Detection of distinct isoform patterns of the β-amyloid precursor protein in human platelets and lymphocytes

NeurobiologyofAging,Vol. 13, pp. 421-434, 1992

0197-4580/92$5.00 + .00 Copyright© 1992PergamonPressLtd.

Printedin the U.S.A.All rightsreserved.

Detection of Distinct Isoform Patterns of the ¢/-Amyloid Precursor Protein in Human Platelets and Lymphocytes M I C H A E L G. S C H L O S S M A C H E R , B E T H L. O S T A S Z E W S K I , L A N N Y I. H E C K E R , * A L E S S A N D R O CELI,* C H R I S T I A N H A A S S , D A V I D C H I N , ~ I V A N L I E B E R B U R G , t B A R B A R A C. F U R I E , * B R U C E F U R I E * A N D D E N N I S J. S E L K O E 2

Center for Neurologic Diseases, Harvard Medical School Brigham and Women "s Hospital, Boston, MA 02115 *Division of Hematology-Oncology, Tufts University School of Medicine, New England Medical Center, Boston, MA 02111 ?Athena Neurosciences, Inc. South San Francisco, CA 94080 Received 8 N o v e m b e r 1991 ; Accepted 17 J a n u a r y 1992 SCHLOSSMACHER, M. G., B. L. OSTASZEWSKI, L. 1. HECKER, A. CELl, C. HAASS, D. CHIN, I. LIEBERBURG, B. C. FURIE. B. FURIE AND D. J. SELKOE. Detectionof distim't isoformpatterns q/the 8-amyloidprecursorprotein in human platelets and lymphocTtes. N EU ROBIOL AGING 13(3) 421-434, 1992.--Cerebral deposition ofthe amyloid 8-protein (ASP), a ~40 residue fragment of the integral membrane protein, ~-amyloid precursor protein 03APP), has been implicated as the probable cause of some cases of familial Alzheimer's disease (AD). The parallels between ASP deposition in AD and the deposition of certain plasma proteins in systemic amyloid diseases has heightened interest in the analysis ofdAPP in circulating cells and plasma. Here, we describe distinct isoform patterns of 8APP in peripheral platelets and lymphocytes. PCR-mediated amplification of mRNA from purified platelets demonstrated the expression of all three major 8APP transcripts (SAPP770,7~.69,).The full-length, ~ 140 kDa form of SAPPT~.770was detected in membranes of resting and activated platelets but very little immature, ~ 122 kDa ~3APP75~jTowas found, suggestinga different processing of SAPP in platelets than that described in a variety of cultured cells and tissues. Platelets stimulated with thrombin, calcium ionophore, or collagen released the soluble, carboxyltruncated form of/3APP(protease nexin-II), but no evidence for the shedding of full-length~3APPassociated with platelet microparticles was found, in contrast to previous reports. As a positive control marker for microparticles, the fibrinogen receptor subunit, GPIlIa, was readily detected in platelet releasates. Resting and activated platelets contained similar amounts of the 10 kDa carboxyl terminal 8APP fragment that is retained in platelet membranes followingthe constitutive cleavageof protease nexin-II. Nonstimulated peripheral B and T lymphocytes contained small amounts of membrane-associated mature and immature/3APP75j.770.The potentially amyloidogenic full-length 8APP molecules present in circulating platelets and lymphocytes but not in microparticles could serve as a source of the microvascular A/3Pdeposited during aging and particularly in AD. Alzheimer's disease Lymphocytes

t~-amyloidprecursor protein

Proteasenexin-II

PROGRESSIVE deposition of the amyloid beta protein (A/3P) in cerebral plaques and meningocerebral blood vessels is a principal feature of Alzheimer's disease (AD; 20,41,63). Highly similar lesions occur in subjects with trisomy 21 (Down's syndrome) and to a lesser extent, in normal aged humans and other primates (19,40,57,64). A/3P is a 39-43 amino acid fragment of a highly conserved, integral membrane protein, the/3-amyloid precursor protein (flAPP). At least five alternatively spliced transcripts of the t/APP gene have been described ( l 1,22,32,33,55,74). Evidence for a probable causative

Platelets

Microparticles

role of/3APP in AD was recently provided by the discovery of at least seven unrelated families with autosomal dominant AD in which affected members have an amino acid substitution in the single transmembrane domain of the protein, i.e., at amino acid 717 in the 770-residue form (/3APP770)(9,21,38,43,46). One defined pathway for the posttranslational processing of /3APP involves a constitutive proteolytic cleavage at residue 687 offlAPP770, causing secretion ofthe ~ 125 kDa extramembranous amino-terminal portion into the extracellular fluid (e.g., cerebrospinal fluid and plasma) and retention of the re-

tCurrent address: Protein Sequencing & Chemistry Facility, University of Missouri/Columbia, Columbia. MO 65211. 2Requests for reprints should be addressed to Dennis J. Selkoe, Center for Neurologic Diseases, Brigham and Women's Hospital. 221 Longwood Avenue, Boston, MA 021151. 421

422

SCHLOSSMACHER ET AL.

maining ~ 10 kDa carboxyl-terminal fragment in the membrane (50,53,61,65,78). The secreted form offlAPP that contains a Kunitz-type protease inhibitor domain has been shown to be identical to protease nexin-II (PN-II) (47,77). Because the constitutive cleavage at residue 687 occurs within the A/3P region (13,68), generation of intact AflP fragments in aging and AD must involve an alternate proteolytic pathway. The cellular origin of those/3APP molecules that give rise to the ACtP deposits is unknown. AD shares several pathophysiolog~cal leatures wtth certa|n systemic amyloid deposition diseases (amyloidoses) known to be of circulating origin (62). For example, the localization of~amyloid fibrils within the media of meningeal blood vessels outside of brain tissue is similar to that of some systemic amyloid deposits (e.g., AA and transthyretin amyloids) that arise from circulating proteins. This observation and others (26,30) make a solely neuronal or glial source of A/3P improbable and have led us to search for potentially amyloidogenic forms of /3APP in extracellular fluids. The known forms of~APP in CSF (50,53,78) and plasma (53) do not contain the complete A~P region and thus could not be amyloidogenic. Such soluble /3APPm/770 fragments are also present in platelet a granules and released upon activation with platelet agonists (69,76). Upon stimulation, platelets are believed to release not only soluble proteins (for review, see 27) but also membranous microparticles that contain proteins such as PADGEM (GMP140/CD62/P-selectin) or GPIIb/lila (a,d3rintegrin) (79). Two recent publications reported that full-length flAPP (FL-{3APP) is released from platelets in membranous microparticles (8,10). This apparent release of the intact precursor into the circulation could have relevance for the fundamental mechanism of fl-amyloidosis in AD. However, the purity of the platelet preparations used in these experiments, the state of activation of the platelets, the possibility that intact platelets contaminate the releasates, and the rigorous identification of the released ¢~APP isoforms have not been fully addressed. Here, we examine human peripheral blood cells as regards the presence and nature of various flAPP isoforms. We report that /3APPm/770 is the predominant form of/~APP in purified platelets, as shown by immunochemical methods and polymerase chain reaction (PCR). After platelets are stimulated with thrombin, calcium ionophore, or collagen, FL-flAPP is found in the cell membrane but cannot be detected in released microparticles, which were identified by the control marker protein, GPIIIa, a subunit of the fibrinogen receptor. Using immunochemical techniques, we confirm the presence of the PN11 portion of ~APP in platelet a granules and also describe the residual ~ l0 kDa membrane bound fragment in platelets. Furthermore, we demonstrate the presence of small amounts of mature and immature FL-BAPPT~,770in nonstimulated circulating B and T lymphocytes. No FL-/3APP was detected in red blood cells or plasma. The presence of FL-flAPP in circulating cells has implications for the pathogenesis of A/3P deposition during normal aging and in AD. METHOD

Plateh,t Preparation Platelets were prepared from whole blood drawn through a 17 or 19 gauge needle and anticoagulated by acid citrate dextrose solution (ACD) (USP Fenwal, Baxter). Donors for all platelet preparations used in this study were healthy volunteers, aged 26 to 47 years, who had been free of medication for I week. Blood was sedimented by centrifugation at room temperature (RT) for 15 rain at 190 g. The resultant platelet-rich

plasma (PRP) was collected. To purify resting, nonactivated platelets from PRP, we used Sepharose 2B chromatography as previously described (4,37,80). Briefly, Econo-Columns (Biorad) with 55 ml bedvolume ofSepharose 2B beads (Pharmacia) were preequilibrated at RT with modified Tyrode's buffer (MTB), which contained 3.7 mM HEPES, 137 mM NaCI, 12 mM NaHCO3, 2.6 mM KCI, 2.4 mM MgCI,, 0.1% dextrose, 0.1% bovine serum albumin and 6 protease inhibitors (5 mM EDTA, 2tsg/ml Aprotinin, 0.5 t~g/ml Leupeptin, 40 ug/ml TLCK, 0.7 ug/ml Pepstatin A and I mM PMSF), unless otherwise noted. In some experiments, acetylsalicylic acid (Sigma) was added to either MTB or donor blood, as described in the Results section. Nine to 10 ml of PRP were loaded per column and up to 10 ml of filtered resting platelets ( ~ 3 × 108/ml) devoid of plasma proteins and nearly all white blood cells (WBC) were collected. ( "ell ('mints

Cell counts were determined on a hemocytometer using the unopette microcollection system (Becton-Dickinson) (6). In all cases, at least four hemocytometer chambers were counted per sample. Platelet preparations for PCR amplification were analyzed for WBC content by six undiluted hemocytometer counts.

P( "R-mediated Ampl(/ication qf Platelet mRNA To study ¢4APP transcripts in platelets completely purified from contaminating WBC, PRP was filtered serially through Sepharose 2B beads and then 5 ml of cotton wool taken from I M U G A R D IG500 columns (Terumo) (e.g., 7) in the absence of protease inhibitors and acetylsalicylic acid. Eluting platelets ( ~ 2 × 108/ml) were pelleted at 2000 g (15 min, RT), frozen, thawed and RNA was isolated by the guanidinium-thiocyanate procedure (39). The first strand cDNA (from 3 ug RNA) was amplified by PCR [58; and manufacturer's protocol for taqpolymerase (Cetus)] using the following oligonucleotides: the antisense primer spans the region between bp 1200 and bp 1219 and the sense primer corresponds to the region between bp 640 and 660 of~APPr~5 (32). Primers were kindly provided by E. Koo. Plateh't l,ysate The total platelet lysate in (shown in Fig. 2A) was prepared from an aliquot of a fresh blood bank platelet unit pelleted at 1000 g X 15 min, resuspended in Tris-buffered saline (50 mM/ 150 mM) and homogenized in 2X Laemmli sample buffer (36). Between two sonication (20s)/boiling ( 10 rain) steps, the sample was stored frozen at -20°C. Prior to electrophoresis, the lysate was clarified by centrifugation at 13,000 g for 3 rain. Plateh,t Reh, asate and Retentate For platelet activation, either 0.25 or 1.0 U/ml thrombin (Sigma), 5 or 10 uM/ml calcium ionophore A23187 (Sigma; reconstituted in dimethyl sulfoxide) or plain MTB was added to gel filtered platelets. After gently inverting the tubes twice, cells were incubated at RT for 15, 25, or 30 min. Immediately following agonist incubation, platelets were sedimented by centrifugation at 1200 g for 15 min at RT, unless otherwise noted. An aliquot of the supernatant (platelet releasate) was i m mediately frozen at - 20°C; the remainder was further spun at 100,000 g for 30 min at 4°C. The resulting pellet was either extracted in 2% Triton X - 100 for further fractionation (see following) or directly solubilized in Laemmli sample buffer and

~-AMYLOID PRECURSOR PROTEIN IN BLOOD CELLS sonicated. The releasates (after storage at -20"C) were subjected to 10% trichloroacetic acid (TCA) precipitation for 60 min at 4", followed by two 15 rain washes in - 2 0 " C ethanol. The TCA-precipitated releasates were then solubilized in Laemmli sample buffer at 20-fold concentration, sonicated for 15 s, and the pH was neutralized. The post-incubation 1200 g pellets (platelet retentates) were resuspended in lysis buffer (50 mM Tris, pH 7.6, with the 6 protease inhibitors) and stored frozen. Retentates were then thawed, sonicated and centrifuged at 100,000 g for 30 min (4"C) to yield a Tris-soluble fraction and an insoluble pellet; the latter was resuspended in lysis buffer, sonicated and extracted for 2 h with 2%Triton X-100 at RT. Triton-soluble material was recovered as the supernatant after a 100,000 g spin (30 min); the Triton-insoluble pellet was resuspended in Laemmli sample buffer. Aliquots of all nonstimulated pre-and post-filtration platelet samples were pelleted at 2 ! 00 g × 15 rain and subjected to the same fractionation steps just described for the post-stimulation retentates, in order to search for differences in the ~APP pattern before and after filtration. All supernatants in this study were assayed for protein concentration according to the dye-binding method (5) before being electrophoresed on SDS/ polyacrylamide minigels (36) and transferred to Immobilon P (Millipore) for Western blotting.

Western Blotting Immunoblots were performed as previously described (53); briefly, acrylamide gels were transferred either overnight or for 2 h at 4"C, followed by blocking in 5% milk (Carnation NonFat)/TBST (tris buffered saline, 0.05% Tween-20) for 45 rain at RT. Primary antiserum incubation (1% bovine serum albumin, Sigma) at indicated dilutions was performed either overnight at 4"C or at RT for 3 h. lmmunoreactive bands were visualized using goat antirabbit or antimouse IgG conjugated with alkaline phosphatase (Promega; secondary antibody incubation at 1:7500 for 30 min at RT; development time 1215 rain). The sensitivity of this immunoblot method for detecting forms of/3APP with the principal antibodies used (at dilutions indicated) is as follows: aC7 (affinity purified) can detect 3.75 pg of recombinantly expressed ~APP~92 695 (48) per gel lane, 200 pg of synthetic peptide t3APP676-695on dot blots, and the FL-t~APP in 5 ug per gel lane of Triton extract of unstimulated, gel-filtered platelets (prepared in the absence of protease inhibitors); aB5 (affinity purified) can detect 16 pg of purified, baculovirus-expressed FL-C~APPT~k(35) per gel lane, 30 pg of purified recombinantly expressed BAPP~-592 (48) in dot blots, and FL-/3APP in 3.75 ug of Triton extract per gel lane of gel-filtered platelets.

Antibodies All antisera were extensively characterized previously by Western blotting of cerebral cortex (49,50,65,72), BAPP-transfected cells (48,65), CSF (50), media (48,53), and plasma (53) and by immuno-cytochemistry (29,49,54,65). Proteins in brain, CSF, and/~APP-transfected cells immunoreactive with these and similar antibodies were previously confirmed to be forms of BAPP by direct sequence analysis ( ! 3,50,72). Rabbit antisera to the/~APP carboxyl terminus were raised against the following synthetic peptides (numbers according to ref. 32 for /~APP69s): otCl (65), aC7, and aC8 (53,54) all to/~APP676_695 [some aliquots of aC7 were affinity purified as described (23)]; aC2 (R37) to/~APP68~-s5 (gift of T. Ishii) (28); aC4 to BAPP~.69s (gift of Y. lhara). Antibodies aB6 to BAPP592-695,

423 aB5 to 8APP,,. s~ and aB3 to flAPP20-~ are affinity-purified rabbit antibodies to bacterially expressed recombinant proteins (48); IG5 is a monoclonal antibody raised against ~APP~-s92 (53). Ri285 is a rabbit antiserum to/3APP527-uo; some aliquots were affinity purified as described (23); an antiserum (R36) to the same antigen was also kindly provided by T. Ishii. aN l and aN3 are rabbit antisera to BAPP4s-62 (54). Monoclonal antibody 22C I I to recombinant full-length 8APP was purchased from Boehringer-Mannheim (78). [On all Western blots, lysates oftransfected 293 cells were electrophoresed as controls for positive BAPP isoform identification]. The monoclonal anti-PADGEM antibody, ACI.2, was previously described (37). SSA6 (kindly provided by S. Shattil) is a protein-A purified monoclonal antibody to the fibrinogen receptor subunit GPIIIa (66). RT 97 (gift ofB. Anderton) is a monoclonal antibody to the neurofilament high molecular weight protein (1). aB- 1 and aT-3 (Coulter Immunology) are fluorescein isothiocyanate (FITC)-conjugated monoclonal antisera to lymphocyte specific surface antigens CD 20 of B-cells and CD 3 ofT-cells, respectively (gift of H. Weiner).

Preparation ~f Lymphocytes Sixty milliliters of ACD-anticoagulated whole blood were centrifuged at 190 g for 15 min (RT), and the PRP was removed. The pellet was diluted with phosphate buffered saline (PBS) to 60 ml and layered over FicolI-Paque solution (Pharmacia) (15). The mononuclear cell layers resulting from 480 g centrifugation × 35 min (RT) were combined, washed in PBS, repelleted at 300 g, and resuspended in RPMI (Sigma)containing 1% (vol/voi) penicillin (5000 U/ml)-streptomycin (5000 ug/ml) mixture (Gibco) and either 5% fetal bovine serum (FBS, Hycione) (Fig. 7), 20% FBS (Fig. 7, transformed cells) or 4% donor serum (Fig. 8); [S. Swaminathan kindly provided the Epstein-Barr virus transformed cell line AM-8 (71 )]. The cells were plated onto Petri dishes and incubated overnight at 37"C (95% 02/5% CO2). Nonadherent lymphocytes were removed with the media and washed in PBS. Cells were either iysed in lysis buffer [also containing 5 uM diisopropyl fluorophosphate (Sigma)] for fractionation studies or were used directly for immunocytochemistry.

Immunofluorescence Microscopy Immunofluorescent staining of platelets was carried out by modifying a previously described protocol (4). Gel filtered platelets were pelleted for 15 min at 2000 g (RT) and resuspended in MTB. Cells were fixed for 60 min in 1% paraformaldehyde at 4"C; remaining aldehyde activity was blocked by washing in 20 mM NH4CI2/PBS. Repelleted platelets (1200 g × 10 min) were suspended in washing buffer (PBS/I% BSA) and allowed to settle for 30 min on poly-L-lysine coated coverslips. Adherent cells were washed × 4 and blocked for 15 min in PBS/50% horse serum. Cells were permeabilized for 10 min in PBS/I% BSA/0.1% Triton and again blocked for 15 min. Following washes, platelets were incubated for I h (RT) with the primary antibody in antibody solution (ABS) containing PBS with 400 mM NaCI/10% goat serum (Cooper-Capell)/ 0.1% Triton. Coverslips were washed in washing buffer containing 750 mM NaCI and incubated with goat a-mouse- or arabbit rhodamine-conjugated IgG in ABS (1:500, BoehringerMannheim) for 30 min. Subsequent washes were performed as after primary incubation. Finally, the coverslips were flushed with deionized water, mounted (90% glycerol/PBS, pH 9.6) and sealed. Staining of lymphocytes was performed by a modification

424

SCHLOSSMACHER ET AL.

of the protocol of Bakke and Dobberstein (2). Isolated cells were resuspended in icecold PBS/2% FBS. For surface antigen labeling, primary antibodies were added to some lymphocytes for 1 h on ice. After washing, cells were fixed for 30 min in 3% paraformaldehyde on ice; excess aldehyde was blocked for 5 min in 100 mM glycine (RT) and cells were resuspended in buffer. For intracellular antigen labeling, some lymphocyte aliquots were fixed, washed, permeabilized and blocked for 10 min in PBS/2% FBS. Cells were then incubated for 60 min and washed in buffer containing 400 mM NaCI. For detection of antibody binding, all lymphocytes were incubated for 30 min on ice with secondary antibody [goat-a-rabbit rhodamine-conjugated IgG in PBS/2% FBS (except aB-I and (,T-3, see above)]. All pelleting steps were carried out by I min spins in a swinging-tube microcentrifuge at 1650 rpm (Fisher, model 59a). Finally, labeled lymphocyte suspensions were cytocentrifuged (SHANDON Cytospin 2) at 1330 rpm × 2 min (RT) onto poly-L-lysine coated microscope slides, coverslipped and sealed. ( Tlrornatographic Fractionation and Analysi.s tffPlasma Initial fractionation of whole plasma from the same donors as for platelets was performed on columns ofagarose-Cibacron Blue F3GA (Affi-Gel Blue, Biorad) as previously described (53). The nonionic detergent, n-octyl-fl-D-glucopyranoside [(Sigma), I%], was present in the 30 mM sodium phosphate buffer, pH 7.0. Affi-Gel Blue fractions were 33 × concentrated by TCA-precipitation and screened by Western blot analysis. Those fractions eluting at 1.5 to 2.0 M NaCI were pooled and following buffer exchange subjected to further purification by loading ~ 2 0 mg onto a Mono-Q HR 5/5 anion exchange column (Pharmacia) in a FPLC chromatography system (Waters 65). TCA-precipitated fractions were again assayed by Western blotting. Pooled fractions containing the bands of interest ( ~ 140 and ~ 170 kDa; see Results), which eluted between 180 and 220 mM NaCI, were loaded onto a preparative 7.5% acrylamide gel (Biorad), electrophoresed and transferred onto Immobilon. Strips containing the immuoreactive ~ 140 and 170 kDa bands were cut out and applied to an ABI Model 470A protein sequencer.

£PI ~i"

NH 2

~B3

a-N

RESULTS lluman P/ateh'ts (.'ontain Full-Length [~APP Polypeptides and mRNAs We initially examined total platelet iysates from blood bank platelet units (see Method Section) by Western blotting with several antibodies previously shown to specifically label flAPP (29,48-50,53,54,65,72) including six distinct antibodies to the carboxyl terminus of the protein (Fig. 1,2A). All antisera ((~B3, (~C 1, (rC2, (~C4. ,C7, ,C8, ,B6) specifically recognized an 138-140 kDa band that comigrated with the full-length/3APP (FL-~APP) overexpressed in human kidney 293 cells transfected with a cDNA encoding flAPP~ (48) (Fig. 2A). To analyze which alternative transcripts of/3APP are expressed in platelets, we performed PCR-mediated amplification o f m R N A isolated from purified platelets. All three major ~APP transcripts, namely ~APP770,/JAPP7~I and/JAPP09~, (Fig. 2B lane 5) were detected in platelet preparations devoid of any contaminating white blood cells (WBC) (see Method and next section). Total RNA isolated from kidney 293 cells transfected with either 8APPT~ or ¢~APP69~cDNAs (48) served as control samples (Fig. 2B lanes 3,4). The PCR-amplified platelet bands comigrated with the corresponding transfected 293 cell products. The identity of the three platelet transcripts was confirmed by Southern blot analysis employing ~APPT~L,770-and ;JAPP6~5-specific cDNA probes under high stringency conditions (data not shown). Protease Nexin-ll But .Not Full-Length I~APP Is Reh,ased I"rmn Activated Platelets Platelets isolated from stored blood bank platelet units ( 14,17) or by differential centrifugation of platelet rich plasma (8.10,44) are known to be partially activated, as determined by the expression of the platelet ~ granule membrane protein, PAIS)GEM, on the cell surface (e.g., 4,17). This fact is important for the rigorous identification of platelet proteins, because WBC are known to bind to activated platelets (31,37). We therefore examined platelets from freshly drawn whole blood that were purified by gel filtration, a procedure which yields cells largely in a resting state (4,37,80). These highly purified

cc H Io o i i

A~P~5COOH

aBS/1 G5

c~C

R12--85 etB6

FIG. 1. Schematic structure of/~APP, showing the immunogens ofantii'x3dicsand the primers used in this study. The horizontal bold line represents/~APP69~ (32). Thc shaded vertical box indicates the membrane spanning domain of eAPP. The white box rcpresents the ~4()-residue amyloid fl protein (Ag/P). The arrow indicates the constitutive cleavage of ~APP in transfected cells (13) that results in a large N-terminal, soluble fragment and a short ~ I() kDa membrane-associated C-terminal fragment. A triangle depicts the site of two alternatively spliced inserts, one of which encodes the Kunitz protease inhibitor-like domain (KPI). Arrowheads represent the primers used tbr PCR. CHO indicates sites of N-glycosylation. Lower lines represent the immunogens of the principal antibodies used in this study (see Method section).

~:/-AMYLOID PRECURSOR PROTEIN IN BLOOD CELLS

425

A M r x 1000

196 .¢1.___138-140

]2s

122

106 71--

44~ 1

234

5 6 7 8 910 11121314

B

Mbp 861

-II

~

7311

615 492 369 246

1

2

3

4

5

i

10 bp ~3API'~e) 520 bp ~APPTSI) 3SObp (~APlq~)

FIG. 2. Detection of t3-amyloidprecursor protein (~APP) and/~APP-transcripts in peripheral platelets. (A) Immunoblot of lysates ( 15 ug/lane) of human kidney 293 cells transfected with a/3APP75t cDNA (lane 1) or total platelets (lanes 2-14) (see Method section) electrophoresed on a 7.5% acrylamide minigel and blotted onto Immobilon (53, also see Method section). Antibodies: lanes I and 2. aB3 (1:400); lanes 3 and 4, aC7 ( 1: 1000); lanes 5 and 6, aC8 ( 1:500); lanes 7 and 8, aC2 ( 1:350): lanes 9 and 10, aB6 (1:350), lanes II and 12, aCl (1:350); lanes 13 and 14, aC4 (1:750). Lanes 3-14: Even numbered lanes show staining after preabsorption ofeach antibody with 5-15 ug of synthetic ¢tAPP676-695peptide per ul antiserum. Arrows indicate estimated molecular masses for the mature FL-/3APP( ~ 138-140 kDa), the protease nexin-II fragment ( ~ 125 kDa) and the immature, N-glycosylatedform of FL-/3APP ( ~ 122 kDa) seen in the transfected cells (lane I). Bands below 106 kDa were only stained by some of the antisera and are not forms ofBAPP. (B) PCR-mediated amplification of platelet derived ¢tAPP transcripts. Amplification products of platelet RNA (lane 5) and RNA of nontransfected (lane 2) and transfected kidney 293 cells (¢~APP69stransfectants, lane 3; ~APPvs~transfectants, lane 4) were separated on a 1.5% agarose gel and visualized by ethidium bromide; Lane 1, molecular size markers. Note the presence of all three major ~3APPcDNA bands in lane 5, corresponding to/3APP770,/~APP75j,and/3APP69~transcripts.

platelets (see Method section) were then studied before and after agonist stimulation to reevaluate reports that platelet agonists such a thrombin and calcium ionophore cause the secretion of the protease nexin It (PN-1I) fragment of BAPP (69,76) and the release of FL-BAPP associated with microparticles (8,10). Using a concentration of thrombin (0.25 U/ml) lower than in previous studies (10) but sumcient to activate platelets (4,37), we observed abundant release ofan ~ 125 kDa

protein detected by an N-terminal BAPP antibody (aB3 to ~3APP20-302)(Fig. 3A lane 2), whereas little spontaneous release occurred in the absence of thrombin (Fig. 3A lane 3). Comigration studies with both cerebrospinal fluid and the media of /3APP75rtransfected cells confirmed that this ~ 125 kDa band migrated with and had the specific immunoreactivities of PNII (data not shown) (see also 69,76). No FL-/3APP was detected in the releasates (Fig. 3A, lanes 2 and 3), whereas it was readily

426

SCHLOSSMACHER ET AL.

A

demonstrated in total platelet lysales (lane 4). Using high thrombin concentrations (3, 5.25, or 10 U/ml) (see 10) under otherwise similar conditions, the resultant releasate contained the ~ 125 kDa PN-II band plus an abundant N-reactive 100110 kDa band but no FL-~APP. The 100-110 kDa band can also be detected in total platelet lysates (Fig. 3A, lane 4) and probably derives from further proteolysis of PN-II by high levels ofthrombin or by intracellular proteases. No lower molecular weight forms of/3APP were detected in the releasates. Microparticles are extracellular lipid vesicles of ~0.05 to 0.6 um diameter which can be separated from the releasates of activated platelets by centrifugal forces greater than 35,000 g (see e.g., 18). Using additional N- and C-terminal/3APP antibodies, we could not detect any FL-/3APP molecules in either total releasates (data not shown) or trichloroacetic acid-concentrated releasates (Fig. 3B) of stimulated platelets or in the analogous supernatants of resting cells. In some experiments. however, we recovered small amounts of both FL-/3APP and PN-II if the platelet releasate was spun at 100,000 g and the pelleted material analyzed, as reported by Bush et al. (8) and Cole et al. (10). To determine the source ofthis trace FL-~APP, we compared the intensity of the FL-~APP band in Western blots of the 100,000 g pellets with the number of intact platelets contaminating the platelet releasates (Fig. 4A). In conventional post-1200 g releasates, counts of ~ 3.8 X l06 intact platelets per ml were associated with the small amounts of FL-BAPP in the 100,000 g pellet (Fig. 4A lanes 3-5). However, if the stimulated platelets were initially sedimented at 5000 g instead of 1200 g, the releasates now contained only ~ 1.4 × 105/ml intact platelets and failed to show a FL-~APP band in the 100,000 g pellet (Fig. 4A lanes 6-8). Western blots of unstimulated platelets confirmed that ~2-2.5 X 106 platelets or 5.0 ug of platelet protein per gel lane are required to detect FLBAPP by immunoblotting with our antibodies. As a positive control to confirm the presence of platelet microparticles in the post- 1200 g and post-5000 g releasates and their 100,000 g pellets, we demonstrated abundant GPIIIa, a subunit of the microparticle-associated fibrinogen receptor (GPllb/llla) (66,67,79), in all of these fractions (Fig. 4B). The detection of more PN-I1 than FL-BAPP in the 100,000 g pellets of the post-1200 g releasates (Fig. 4A lanes 3-5) and the similarity of this pattern to that of intact platelets (Fig. 4A lane 2) support the conclusion that the trace FL-BAPP found in the post-1200 g releasate is likely to derive not from microparticles but from contamination of the releasate by intact platelets. Our failure to detect more FL-BAPP in the pelleted releasates of activated platelets than resting platelets (Fig. 4A. compare lanes 3 and 4), while showing such an increase with the control microparticle marker, GPIIla (Fig. 4B, compare lanes 2 and 3, 4 and 5), argues further against the shedding of FL-~APP associated with microparticles from stimulated platelets. Similar results to those described here were observed in the releasates of platelets stimulated with 3, 5.25 or l0 U thrombin/ml (data not shown). Moreover, both collagen-activated platelets and platelets enriched by differential centrifugation (as in 8,10) provided similar findings. Figure 5 analyzes the r e t e n t a t e s (i.e., the post- 1200 g pellets) of the same stimulated platelets that generated the releasates shown in Fig. 4A. Sequential fractionation of these retentates demonstrated more residual PN-II in the cytosolic (not shown) and Triton-soluble fractions (Fig. 5, lane 2) of nonactivated platelets than in the corresponding fractions of activated platelets (Fig. 5, lanes 3 and 4), as expected (see also, 8). This finding indicates that some PN-II is retained inside the platelet following c~ granule release. Moreover, the FL-BAPP protein is de-

M r x 1000 1

2

3

4

196 140 122 ÷ 125 105

106

71--

LL

NC NC NC

NC

Mr x 1000

!

!

196 --

4--125

106 -I

7144--

i :

i

12345 FIG. 3. lmmunoblots of concentrated post-1200 g-releasates of gel tiltered platelets following thrombin stimulation. Cells were prepared in the presence of 0.5 uM acetylsalicylic acid without protease inhibitors. Samples were electrophoresed on 7.5% acrylamide minigels and blotted onto Immobilon. (A) Lanes: I, transfected 29375~ cells (30 ug): 2 and 3, releasates (4 ~.l/lane) after incubating platelets (30 min) with either 0.25 U/ml thrombin (lane 2) or modified Tyrode's buffer alone (lane 3); 4, Triton extract (25 u.g) of unstimulated platelets (derived from ~ 1.5 × 10 7 platelets). Each lane was cut in half and probed with an N-terminal BAPP antibody (N) (aB3 at I: 300) and C-terminal antibodies (C) (mixture ofaC2 and aC7, each at 1: 1000). Arrows indicate molecular masses of~APP proteins as described in Fig. 2; arrow at 105 points to the ~ 100-110 kDa soluble N-reactive band (see text). The nature of the N-reactive smear above the ~ 125 kDa PN-II protein observed in platelet samples stained with N-terminal antisera is not known but can be seen with and without agonist stimulation and in the absence or presence of protease inhibitors in modified Tyrode's buffer. Bands below ~ 100 kDa (including the ~71 kDa band) are nonspecific. (B) Thrombin-stimulated ( 15 rain) releasate (3 ul/lane) of the same platelet preparation studied in Fig. 3A probed with N- and Cterminal ~APP antisera; Lanes: 1, ~B3 (1:300); 2, 22CI I (5t~g/ml); 3, c~C7 ( 1:750)" 4, a,C8 ( 1:500); 5, aC2 (1:350). Note the presence of the C-truncated PN-II protein ( 125 kDa) but the absence ofFL-BAPP molecules ( 140 kDa) in the releasate.

/3-AMYLOID PRECURSOR PROTEIN IN BLOOD CELLS

427

A

M r x 1000 111 - 71--

43~

1

2

3

4

5

6

7

8 140 "~- 122 ÷ 125

i

NC NC NC NC NC NC NC NC

+

Mr x 1000

--

+

i i

111 m

95 71--

i

I

!

!

43--

i 1

2

3

4

FIG. 4. Immunoblots of microparticle preparations probed with antisera to BAPP (A) or to the fibrinogen receptor subunit, GPIlla, (B). Plateletswere prepared by gel filtration in the presence of protease inhibitors but without acetylsalicylic acid. (A) Comparable aliquots (average3.4 X 108 cells/ml) of platelets from one donor were incubated (25 rain) with either modified Tyrode's buffer alone (lanes 3,6), calcium ionophore ( 10uM/ml) (lanes 4,7), or thrombin ( I U/ml) (lanes 5,8). Cells were then separated from the releasatesby sedimenting at either 1200 g (lanes 3-5) or 5000 g (lanes 6-8). An aliquot of each supernatant was counted for platelets; the rest was spun at 100,000 g to pellet micropanicles and any remaining platelets. The high-speedpellets were solubilized in 2X sample buffer and half of each sample was loaded per lane (5%-20% acrylamide minigel). Lane 1, homogenate of ~APP75t-transfectedcells (2"/~,g); lane 2, Triton extract (32 ug) of total platelets prior to incubation (derived from ~2.4 X 108 platelets). Each lane was cut in half and probed with an N-terminal antibody (aBL 1:2000) and a C-terminal antibody (a~nity purified ,~C7, 1:40). Note that in the post-1200 g supernat~nts (lanes 3-5), that contained an averageof 3.8 X l & platelets/ml, the high-speedpellets show small amounts of FL-BAPP (140 kDa) and abundant PN-ll (125 kDa), whereas the high-speed pellets of the post-5000 g supernatants (lanes 6-8), containing an averageof 1.4 × I 0 5 platelets/ml, show no FL-~APP and little PN-II. (B) Approximately 3.2 X 108 platelets/ml were incubated (25 rain) with either modified Tyrode's buffer alone ( - ) (lanes 2,4) or calcium ionophore( + ) (lanes 3,5). Releasateswere separated from the platelets by centrifugation at 5OOOg; 2.5 ml of each supernatant was concentrated 6-fold in microconcentrators [Centricon- 10 (Amicon): 50 min, 1500 g, 4°C], and 60 ug were loaded (lanes 2,3); the remaining 3 ml were spun at 100,000g. The highspeedpellets were solubilized in 2 X Laemmli sample buffer and a third was loaded (lanes 4,5). Lane 1, Triton extract (30 ug) of gel filtered platelets prior to incubation. "/.5%acrylamide gel, nonreducing conditions. The blot was probed with antibody (SSA6 ( 15 ,g/ml) to the fibrinogen receptor (GPllla). Note the 92-95 kDa microparticle-associated antigen GPIIIa in platelet extracts and activated platelet releasateand its enrichment in the 100,000g pellet thereof. Separation of cells from their releasatesby a 1200 g spin produced the same immunore~ctive pattern.

428

SCHLOSSMACHER E-I A[,.

Mr x 1000

1

206--~

2 ,!

.

3 4 ~ [

5 140 122 + 125

43--

" --4:

--

--

m

28--

;l-am -i=

,

-

¢--- 1 0

CN CN CN CN CN FIG. 5. lmmunoblot of the platelet pellets (retentates) of the same platelet preparation shown in Fig. 4A. Following agonist incubation and 1200 g sedimentation, platelet retentates were fraetionated (see Method section) and electrophoresed on a 5%20% acrylamide minigel. Lane 1, Triton extract of total platelets prior to agonist incubation (32 ug). Lanes 2-4: Triton extracts of retentates (27 ug) following incubation with either buffer (lane 2) (3.8 × 10s platelets/ml); calcium ionophore (lane 3) (3.8 X 108/ml); or thrombin (lane 4) (2.7 × 108/ml). Lane 5, #APP75rtransfected 293 cells (27 ug). Each lane was cut in half and probed with a C-terminal antibody (C) (~C'7, I: 1000)or an N-terminal antibody (N) (aB5, 1:2000). Arrows at 140 and 10 denote the FL-/3APPand 10 kDa fragment, respeetively, found in all platelet sampies as well as in the transfected cells. Note thai lane 4 contains retentates from a lower number of starting platelets than lanes 2 and 3. Other immunoreaetive bands below 111 kDa (~70, ~50, ~35, and ~22 kDa) did not stain with all of our C-terminal antisera and do not represent/3APP fragments; they probably result from crossreactivity of the non-al~nity purified ~ ' 7 with other proteins.

tected in comparable amounts in the retentates of activated (Fig. 5, lanes 3 and 4) and resting (lane 2) platelets; this ~ 140 kDa band comigrates with the mature, N- plus O-glycosylated precursor protein in ~APP75rtransfected cells (lane 5). In the course of examining platelet retentates, we found that inclusion of protease inhibitors helped prevent cell loss due to adherence to surfaces as well as aggregation-induced activation of sedimented platelets and the resultant partial proteolysis of FL¢~APP. However, we found no qualitative differences in 13APP patterns between platelet/micropartiele preparations studied in the presence or the absence of protease inhibitors.

Platelet Membrane Fractions Also Contain the ~ I0 kDa Carboxyl- Terminal 13APPFragment The origin of the abundant PN-11 molecules stored in platelet a granules has not yet been rigorously established. Some proteins, such as fibrinogen, are believed to be taken up by platelets from plasma whereas others, like yon Willebrand's factor or epidermal growth factor, are synthesized in the megakaryocyte and stored in platelet a granules (3,25,45). In this regard, the Triton extracts of the retentates of both stimulated and nonstimulated platelets (Fig. 5) contained the stable 1013 kDa C-terminal fragment of/3APP that has been described in numerous cells and tissues expressing ~APP (48,54,65). Characterization with six C-terminal/3APP antibodies before and after peptide absorption (data not shown) specifically con-

firmed this ~ 10 kDa platelet protein to be the membrane-retained fragment of/3APP that results from the constitutive cleavage of FL-~APP that generates PN-II (13). We detected similar amounts of the ~ 10 kDa band in the membrane extracts of platelets before and after gel filtration as well as in the retentates of sti m ulated versus non st i m u lated platelets (Fig. 5, lanes 2-4, and legend). These data indicate that both, the ~ 10 kDa protein and PN-II are not created during platelet isolation or agonist stimulation but rather are endogenous constituents of circulating platelets. They further suggest thal constitutive cleavage of FL-/3APP to produce PN-II occurs within megakaryocytes or platelets. Membrane extracts of platelet retentates only occasionally contained trace amounts of an ~ 122 kDa N- and C-reactive protein (not shown). This band is likely to represent an immature. N-glycosylated /3APPTs~/770 isoform which has previously been detected in many cells and tissues expressing ¢~APP (48,54,78). Our data consistently showed that most FL-BAPP occurs in platelets as the mature, N- plus O-glycosylated /3APPT~770 isoform, which migrates at ~ 140 kDa (Fig. 2A): this is in agreement with the data ofGardella et al. (16).

hnm unofluorescence Microscopy C'onJirms the Presence uf N- and C-Terminal [5APPEpitopes in Platelets To localize ¢~APP epitopes within the platelet, we performed immunofluore~ent labeling of gel filtered platelets. When nonstimulated, Triton-permeabilized platelets were examined for/~APP using ~BS, an affinity purified antiserum to recombinant/:~APP444 ~9.,,strong labeling of the platelet cytoplasm was observed (Fig. 6b). The pattern of reactivity was punctate and granular, but slightly less extensive than the staining obtained with an antibody to PADGEM protein (Fig. 6A), a well-characterized ~ granule membrane constituent (e.g., 4,37). When platelets were incubated with c~B6, an affinity purified antiserum to recombinant 8APP~9.,_ 695,the fluorescent signal also appeared punctate, was specific (compare Figs. 6C and D) but was weaker than that obtained with ~B5 (Fig. 6B). Immunolabeling with Mab 22CI1 (to an N-terminal ¢4APP epitope) or affinity purified aC7 (to the C-terminus) produced staining patterns similar to those of c~B5 and aB6, respectively (data not shown). Figs. 6E and F demonstrate no reactivity of platelets with an irrelevant antibody or with secondary antiserum alone. We interpret the ~APP N- and Creactivity seen in Figs. 6B and C to principally represent P N It and the ~ 10 kDa protein, which are both abundantly present in platelets (Fig. 5). The specificity of the/3APP antibodies used in these experiments has also been previously demonstrated by immunocytochemistry oftransfected and nontransfected kidney 293 cells (65). Due to the low fluorescent signal and imperfect subcellular resolution, attempts to detect any redistribution and potential externalization of N- and C-terminal 8APP epitopes following platelet activation have been inconclusive to date.

L ymphocytes Contain Small .4mounts ~f Mature and hnmature FL-BAPP lmmunoblots of membrane extracts of nonadherent peripheral blood mononuclear leukocytes (iymphocytes), comprising both B and T cells, showed small amounts of mature, 140 kDa and immature, ~ 122 kDa FL-~APP7~I/770(Fig. 7, lane 3). Lymphocytes showed less of the ~ 125 kDa PN-II (Fig. 7, lane 3) and the ~ 10 kDa fragments (data not shown) than platelets, suggesting differential utilization of catabolic path-

¢4-AMYLOID PRECURSOR PROTEIN IN BLOOD CELLS

429

M r x 10 an

4t

1

2

3

4

206 Q

Ip

Q ~r

I$

o

~

140 122+125

j

111

t¢

71

O ,a,

ag

•

NC

NC NC NC

FIG. 7. Immunoblot of peripheral lymphocytes and cultured transformed lymphoblasts probed with BAPP antibodies. A 50/0-200/0minigel contains: lane I, homogenate (30 ttg) of BAPPvsrtransfected 293 cells; lane 2, homogenate (60 ug) of nontransfected 293 cells; lane 3, Triton extract (60 ug) of nonadherent peripheral lymphocytes; lane 4, Triton extract (35 ug) of EBV-transformed lymphoblasts. Each lane was cut in half and probed with the N-terminal antibody, aB5 (lanes 1-3, 1:2000; lane 4, I : 1500) and the C-terminal antibody aC7 (lanes 1-3, affinity purified aC7, 1:40; lane 4, antiserum aC7, 1:700). The cell line AM-8 (lane 4) is derived from Epstein Barr virus transformation of normal human B cells from an EBV-seronegative donor (71 ). Cells were cultured and extracted as described in the Method section.

confirmed these findings (data not shown). Lymphocytes probed with an irrelevant control antibody (Fig. 8G) or secondary antibody alone (Fig. 8H) showed no reactivity.

Lack of Detection of FL-13APP in Human Plasma

FIG. 6. Immunofluorescence microscopy of peripheral resting platelets. Gel filtered platelets were prepared for indirect immunofluorescent staining with #APP and control antibodies as described in the Method section. A: monoclonal anti-PADGEM antibody ACI.2 ( I: 1.25); B-D: anti-/~APP antibodies, B: aB5 ( 1:200); C: aB6 ( l: 100); D: aB6 preabsorbed with 5 zg each of synthetic peptides/~APP676 695 and ~APP649-664; E: monoclonal neurofilament antibody RT 97 { 1:200); F: no primary antiserum. Photographs were taken through a Zeiss Axioskop equipped with a 63 × Plan-Neofluar lens. Bar equals 5um. ways for t3APP in platelets and lymphocytes. Triton-extracts of T cells isolated from peripheral blood (data not shown) and of cultured Epstein Barr virus-transformed B cells (Fig. 7, lane 4) also contained both FL-/3APP isoforms. Immunofluorescent surface labeling of living peripheral B and T cells with the control antibodies, aB-I (against pan-B cell surface antigen, CD 20) and aT-3 (against pan-T-cell surface antigen, CD 3) identified both types of cells in peripheral lymphocyte preparations; the major cell-type was the T cell (Fig. 8A,B). aB5 (to ~APP4,u-~92) failed to label the plasma membranes of living, nonpermeabilized T and B cells (Fig. 8C,D). Staining for internal/~APP on fixed, permeabilized cells revealed perinuclear aB5 reactivity in all lymphocytes; the staining was accentuated near the hilus of the kidney-shaped nucleus (Fig. 8E). Reaction with affinity purified aC7 (Fig. 8F) produced a weaker but similar signal. Incubation with other ~APP antibodies (aB6; 22C1 I; at~nity-purified a~APP~27_ ~ )

Following the detection of PN-II in human plasma (53), we continued our search for potentially amyloidogenic forms of /3APP that contain the intact A/3P region in plasma. Because such molecules might be expected to be associated with lipid, we performed Affagel Blue colu m n chromatography of plasma (53) in the presence ofthe nonionic detergent, n-octyl t3-D-glucopyranoside. Pooled, TCA-precipitated column fractions were screened for/3APP-immunoreactive proteins on Western blots. In the 1.8 M NaCI eluate, two bands migrating on immunobots at ~ 140 and ~ i 70 kDa were specifically labeled by ~APP antibodies, aC4, aC7 and a~APP~27-~ (Fig. 9A). In decreasing levels of intensity and specificity, aC8, aB5, and a corresponding monoclonal antibody [ lG5 (53)], plus three N-terminal BAPP antisera laB3, aN~ and aN3 (54)], also labeled these bands (data not shown) (60). However, other antisera, e.g., aB6, aC I, aC2, and a ~ A P P 649 664did not recognize these bands, raising the possibility that these were distinct plasma proteins that shared only some epitopes with ~APP. Chromatography of the Affigel Blue eluates containing these proteins on a Mono-Q anion exchange column followed by preparative SDS-PAGE and microsequencing revealed both the ~ 140 and 170 kDa polypeptides to be forms of human complement factor H (56) (Fig. 9B). The ~ 170 kDa protein began at residue l of factor H and appeared to be the full-length molecule whereas the ~ 140 kDa protein began at residue 324. Sequence comparison of BAPP and complement factor H (l 2) revealed homologies between the two proteins at the antigenic sequences for the N-terminal, midregion and C-terminal ~APP antibodies that crossreacted with complement factor H on our Western blots (Fig. 9C). No evidence of specific carboxyl terminal containing ~APP isoforms has yet been found in plasma.

430

m

SCHLOSSMACHER ET AL.

:El

l]

!I1

FIG. 8. Immunofluorescence microscopy of peripheral lymphocytes. Peripheral lymphocytes were prepared for immunofluorescent labeling with/JAPP and control antibodies as described in the Method section. Panels A-C show surface staining of living cells: A: T-lymphocyte marker aT-3 (1:20): B: B-lymphocyte marker aB-I (1:20): C: aB5 (¢APP,u~-592)(1:40). D: phase microscopy of same field as in C. Panels E-H show internal staining of fixed cells: E: aB5 (BAPP~ 592)( 1:40)F: al~nity purified aC7 ( 1:4); G: monoclonal neurofilament antibody. RT-97 ( I: 100): H: no primary antibody. Bar equals 10 urn.

DISCUSSION

The recognition that multifocal cerebral deposits of A~P are a characteristic feature of aging in a variety of mammals including humans, and the implication of accelerated At3P deposition as a seminal event in some cases of familial AD

underscore the need to define the expression and processing of ~APP in many cell types. In particular, the parallels between #-amyloidosis in AD and the tissue deposition of fragments of circulating proteins in certain systemic amyloidoses [e.g., transthyretin in familial amyloidotic polyneuropathy; serum AA protein in secondary amyioidosis; cystatin C in hereditary cerebral hemorrhage with amyloidosis (HCHWA)-lcelandic type] recommend the study of BAPP in both blood cells and plasma. To date, no amyloidogenic/~APP fragment (i.e., one containing intact ASP) has been identified in human CSF and plasma. Two recent reports identifying apparently full-length ~APP in the releasates of platelets (8, I 0) provided a potentially important hypothetical mechanism for the delivery of/3APP molecules from peripheral blood to the cerebral vasculature. We have characterized the isoforms of BAPP in platelets and lymphocytes based on recognition by multiple #APP antibodies and co-migration with ~APPT~ in cDNA-transfected 293 cells. By directly applying several sensitive C- and N-terminal antibodies to platelet releasates and their high speed pellets, we found no evidence for the release of FL-#APP in membranous microparticles from resting (8) or stimulated (8,10) platelets. Instead, we found that the intact precursor resides within platelet membranes and that the occasional observation of FL-/~APP molecules in platelet releasates is due to the presence of intact platelets which are not quantitatively sedimented by centrifugation at 1200 g or 1500 g, as used in two prior studies (8,10). Bush et at. inferred that FL-~APP was released in microparticles by showing that platelet retentates (i.e., the post 1500 g pellet) contained less aC-reactive ¢~APP after stimulation than before and that the high speed supernatant of the releasate contained no c~C-reactive/~APP. However, they did not directly probe immunoblots of either the platelet releasate or its high speed pellet with a C-terminal antibody. In contrast, we searched directly for C-reactive FL4JAPP in platelet releasates after 5000 g centrifugation, a force which more completely separates intact platelets from microparticles. We were unable to detect FL4JAPP in such releasates or their high speed pellets using N- and C-terminal antibodies (aB5 and a,C7) shown to have high sensitivity for FL-t~APP (see Method). In direct comparison to the experiments of Cole et al., we have stimulated platelets with high doses of thrombin (10 U/ml); in addition, we have employed lower, less nonphysiological concentrations of thrombin (0.25 U/ml) that have been shown to activate platelets and release a granule contents (4,37). In both cases, we observed abundant PN-1I secretion without demonstrable release of FL-/~APP associated with microparticles; the latter were readily detected in our releasates using the control marker protein, GPIIIa. In the high thrombin experiments, we also detected C-terminal-truncated ~APP forms of ~ 105 kDa in releasates (see also 8); these appear to result from secondary proteolysis of released PN-II by abundant thrombin, an artifact that is largely eliminated by using low thrombin concentrations. At the protein level, our characterization of~APP polypeptides in resting platelets reveals that the full-length precursor is almost solely present as the mature, ~ 140 kDa ~APP75~m0species, with only trace amounts ofthe immature, ~ 122 kDa isoform detectable. This protein pattern is in contrast to previously characterized cells and tissues--such as nontransfected and transfected HeLa (78) or kidney 293 (48,65) cells, cultured rat astrocytes and microglia (23), and human, monkey, and rodent brain and nonneural tissues (54,65)--and might be related to the vestigial rate of new protein synthesis in platelets. Such low level de novo/~APP synthesis is indeed suggested by preliminary data we have obtained from 35S-methionine met-

431

~ - A M Y L O I D PRECURSOR PROTEIN IN BLOOD CELLS

A

s

M r x 1000

1

l i r a

[m

196 - - ! 1 1 1 1 ~ . .

XPYFPVAVGKYYSYYXDEEF

Compl. factor H

RPYFPVAVGKYYSYYCDEHF

llllllllllllll

" 170

~ - - 140

106 - 71--

324

170 kDa protein

1 15 XXXNELPPXXNTEIL

IIIII

EDCNELPPRRNTEIL I 15

45 62 MNVQNGIG~SDPSGTKT-C

~APP(45_62 )

II Compl. fact. H

II

IIl

AKCLGEKWSHPPSCIKTDC 953 971 527 540 TTVELLPVNGEFSL

~APP(527_540)

Ill Compl. fact. H

l

PPPELLNGNVKEKT 614 627 666 695 AVTPEERIILSIOIQQNGYENPTYKFFEQMQN

~APP(666_695) fact.

IIl 343

IIIII Compl. factor H

1234

Compl.

20

140 kDa protein

H

I I

Ill

Ill

II

ISHGWAI~ISDSYQYG-EEVTYKCFEGFGI 919 947

FIG. 9. Immunoblot and sequence analysis of plasma proteins reactive with/~APP antisera. (A) Immunoblot of fractionated human plasma probed with BAPP antibodies. Lanes l --4: ~ 1.5 M NaCI eluate (8 ul/lane) from Atfigel-blue chromatography of detergent-treated plasma (see Method section); lanes l and 2, stained with a mixture ofaC7 and aC4 ( l: 1000 each); lanes 3 and 4, aR36 (1:300). Antisera used on lanes 2 and 4 were preabsorbed with 45 or 30 ug per ul of synthetic peptides BAPP676-695or BAPPs27-5~ respectively. Both antisera specifically label two plasma proteins of ~ 170 and ~ 140 kDa. 5%-20% acrylamide minigel. (B) Amino acid sequence analysis of the purified ~ 140 and ~ 170 kDa protein reveals essentially complete homology with human complement factor H. Vertical bars indicate homologous residues between our microsequencing data and the factor H sequence from the University of Wisconsin-Sequence Analysis-Databank(12). X, residue not interpretable. (C) Best-score sequence matching (12) of the three relevant ~APP synthetic peptides with human complement factor H reveals 29%-37% homology as indicated by vertical bars. Sequence numbering in Fig. 9B and C is according to BAPP695(32). and as reported by Ripoche et al. for complement factor H (56).

abolically labeled platelets (M. G. S. and D. J. S., unpublished observation). Our results, taken together with the quantitative storage of the PN-II fragment in a granules and the abundance of the resultant ~ l0 kDa C-terminal protein in platelet membranes, suggest that megakaryocytes and platelets utilize a pathway ofintracellular processing o f ~ A P P that results in high levels of the constitutive cleavage described by Esch and colleagues (13). The fact that cultured glial cells and neurons show the opposite pattern--little secretion of PN-II and formation of the ~ l0 kDa fragment and prominent intracellular insertion of FL-~APP (23)--suggests that different cells process ~APP by routes that favor either a principally extracellular or intracellular fate for the molecule. In platelets in particular, it

will be of interest to assess the subcellular localization of FL~APP and whether it is externalized into the plasma membrane during activation, as has been shown for P A D G E M (4,37). B and T lymphocytes isolated from peripheral blood as well as tranformed B lymphoblasts showed expression o f mostly mature and immature FL-~APP75wT0 forms, with less generation of PN-II and the ~ l0 kDa fragment than in platelets. Our Western blot and immunocytochemical data point to the presence of principally full-length forms of the precursor in lymphocytes and a probable internal membrane-associated localization, similar to findings we obtained in cultured astrocytes and microglia (23). Moenning et al. recently described the p r y -

432

S C H L O S S M A C H E R ET AL.

tohemagglutinin-induced transcription and translation of BAPP in lymphocytes (42). Our data indicate that, constitutive 13APP synthesis occurs at low levels in nonstimulated peripheral lymphocytes. We have not detected either full-length or truncated forms of BAPP in peripheral red blood cells. Likewise, neither we nor others have yet identified FL-/~APP or A/3P-containing fragments in normal plasma. We undertook analyses ofplatelet BAPP because of both the vascular predilection of m a n y A~P deposits in aged animals and h u m a n s and the reports of platelet m e m b r a n e abnormalities in some patients with AD. Z u b e n k o and colleagues described increased platelet m e m b r a n e fluidity as a possible marker for some forms of familial AD with early onset of symptoms (83-85). Several laboratories have reported additional alterations in platelets of AD patients compared to control subjects (52 and references therein). Whether these various findings of platelet abnormalities in AD relate in any way to changes in the m e m b r a n e - s p a n n i n g ~APP molecule in platelets remain to be seen. The deposition of A~P in CNS arteriovenous malformations of aged h u m a n s (26) and its often marked accumulation in meningeal and cerebral capillaries, arterioles, and venules in AD D o w n ' s syndrome, H C H W A - D u t c h type, and sporadic cerebral amyioid angiopathy suggest the involvement of endothelial or other vascular c o m p o n e n t s in B-amyloidosis. As in certain systemic amyloidoses (34,82), ultrastructural studies in AD brain reveal ~-amyloid fibrils in the basement m e m b r a n e of arterial walls and capillaries (81 ). It is not yet clear whether reported structural abnormalities in capillaries and other microvessels in postmortem A D brain (51,59) represent early or late events in the pathogenesis of the disease and how they relate to basement membrane/3-amyloidosis. T a m a o k a et al. recently identified a stable, potentially amyloidogenic ~ 2 2

kDa fragment of C4APP containing the intact AC4Pselectively in brain microvessels (73). These and other data raise the question of whether circulating platelets that interact with vessel walls in aged and AD brains could serve as a sourcc of FLI~APP (or fragments thereo0 that is deposited in the subendothelial matrix, where extracellular proteolysis might lead to the deposition of A~P fibrils in the basement membrane. Such a mechanism could relate to the recent description of a b u n d a n t A/~P deposits in the brains of 15 of 20 n o n d e m e n t e d patients with critical coronary artery disease but in only 2 of 16 nondemented subjects without heart disease (70). A relation between platelet function and the complications of coronary artery disease has been widely postulated. For example, Trip and colleagues (75) recently described a positive association between spontaneous platelet aggregation in vitro and both mortality and cardiac events in survivors of myocardial infarction. We arc now initiating experiments to determine whether ~APP or amyloidogenic fragments thereof show altered intracellular processing or extracellular release in the platelets of familial AD versus normal subjects. ,'~CKN(.)WLED(it5MENrs We thank S. Greenhalgh and H. Yamaguchi for helpful suggestions on immunocytochemistry and M. Ehrhardt for assistance in protein microsequencing. We are grateful to W. Dzik, S. Fiore, and G. Gilbert for advice on platelct purification, mononuclear cell preparations and microparticle preparations. We thank M. Podlisny and A. Mammen for numerous discussions and critical review of the manuscript. M. Taleas and N. Boucher assisted in preparing the man uscript. This work was supported by N 1H grants AG0791 I (LEA D Award) and AGO6173 and the Broad Family Foundation (DJS), by NIH Institutional NRSA T32HL07437 (I,H), and NIH grant HI,42443 (BF.BCF) and by a fellowship from Boehringer Ingelheim (('H).

REFERENCES 1. Anderton. B. H.: Breinburg, D.; Downes, M. J.; Green, P. J.: Tomlinson. B. E.: Ulrich, J.; Wood, J. N.: Kahn, J. Monoclonal antibodies show that neurofibrillary tangles and neurofilaments share antigenic determinants. Nature 298:84-86: 1982. 2. Bakke, O.; Doberstein, B. MHC Class II-associated invarient chain contains a sorting signal for endosomal compartments. Cell 63:707-716: 1990. 3. Ben-Ezra, J.; Sheibani. K.; Hwang, D. L.; Lev-Ran, A. Megakaryocyte synthesis is the source of EGF in human platelets. Am. J. Pathol. 137:755-759; 1990. 4. Berman, C. L.; Yeo, E. L.; WenceI-Drake, J. D.; Furie, B. C.: Ginsberg, M. H.; Furie, B. A platelet ~ granule membrane protein that is associated with the plasma membrane after activation. J. Clin. Invest. 78:130-137:1986. 5. Bradford, M. A rapid and sensitive method for the quantitation of ~g quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254: 1986. 6. Brecher, G.: Cronkite, E. P. Morphology and enumeration of human blood platelets. J. Appl. Physiol. 3:365; 1958. 7. Brubaker, D. B.; Romine, C. M. The in vitro evaluation of two filters (Erypur and Imugard IG500) for whole blood cell-poor platelet concentration. Transfusion 28:383-385; 1988. 8. Bush, A. I.; Martins, R. N.; Rumble, B.; Moir, R.; Fuller, S.: Milward, E.: Currie, J.; Ames, D.; Weidemann, A.; Fischer, P.: Multhaup, G.; Beyreuther, K.; Masters, C. L. The amyloid precursor protein of Alzheimer's disease is released by human platelets. J. Biol. Chem. 265:15977-15983; 1990. 9. Chartier-Harlin, M. C.; Crawford, F.; Houlden, H.; Warren, A.; Hughes, D.; Fidani, L.; Goate, A.: Rossor, M.: Roques, P.; Ilardy, J.: Mullan, M. Early-onset Alzheimer's disease caused by mutations at codon 717 of the t~-amyloid precursor protein gene. Nature, 353:844-846:1991.

10. Cole, (i. M.: Galasko. D.: Shapiro, I. P.: Saitoh, T. Stimulated platelets release amyloid 8-protein precursor. Biochem. Biophys. Res. Commun. 170:288-295; 1990. 11. deSauvage, F.: Octave. J.-N. A novel mRNA of the A4 amyloid precursor gene coding for a possibly secreted protein. Science 245:651-653: 1989. 12. Devereux, J." Haeberli, P.: Smithies. O. A comprehensive set of sequence analysis programs for the Vax. Nucleic Acids Res. 12:387-395: 1984. 13. Esch. F. S.: Keim, P. S.; Beattie. E. C.; Blacher, R. W.: Culwell, A. R.: Oltersdorf, T.; McClure. D.: Ward. P. J. Cleavage ofamyloid peptide during constitutive processing of its precursor. Science 248:1122-1124: 1990. 14. Fijnheer, R.: Modderan, P. W.; Veldman, H.; Ouwehand, W. H.; Nieuwenhuis, H. K.: Roos, D.: de Korte, D. Detection of platelet activation with monoclonal antibodies and flow cytometry. Transfusion 30:20-26: 1990. 15. Fotino. M.: Merson, E. J.: Allen, F. t t. Micromethod lbr rapid separation of lymphocytes from peripheral blood. Ann. Clin. Lab. Sci. 1:131-133:. 1971. 16. Gardella, J. E.: Ghiso, J.; Gorgone, G. A.; Marratta, D.: Kaplan, A. P.; Frangione, B.: Gorevic, P. D. Intact Alzheimer amyloid precursor protein (APP) is present in platelet membranes, and is encoded by platelet mRNA. Biochem. Biophys. Res. Commun. 173:1292-1298; 1991. 17. George, J. N.; Pickett. E. B.; Heinz, R. Platelet membrane glycoprotein changes during the preparation and storage of plateletconcentrates. Transfusion 28:123-126:1988. 18. George, J. N.- Thoi, L. L.: McManus, I,. M.: Reimann. T. A. Isolation of human platelet membrane microparticles from plasma and serum. Blood 60:834-840; 1982. lt~. Giaccone, G.: Tagliavini, F.: Linoli, G.: Bouras, C." Frigerior, I,.:

~ - A M Y L O I D P R E C U R S O R P R O T E I N IN B L O O D C E L L S

20.

21.

22.

23.

24. 25.

26.

27. 28.

29.

30. 3 I.

32.

33. 34.

35.

36. 37.

Frangione, B.; Bugiani, O. Extracellular preamyloid deposits precede neuritic degeneration and senile plaques. Neurosei. Left. 97:232-238; 1989. Glenner, G. G.; Wong, C. W. Alzheimer's disease: Initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem. Biophys. Res. Commun. 120:885890; 1984. Goale, A.: Chartier-Harlin, M.-C.: Mullah, M.; Brown, J.; Crawford, F.; Fidani, L.; Giuffra, L.; Haynes, A.; Irving, N.; James, L.; Mant, R.; Newton, P.; Rooke, K.; Roques, P.; Talbot, C.; PericakVance, M.; Roses, A.: Williamson, R.; Rossor, M.; Owen, M.; Hardy, J. Segregation ofa missense mutation in the amyloid precursor protein ge ne with familial Alzheimer's disease. Nature 349:704-706; 1991. Golde, T. E.; Estus, S.; Usiak, M.; Younkin, L. H.; Younkin, S. G. Polymerase chain reaction amplification orb amyloid protein precursor mRNAs: Identification of a novel alternatively spliced form and analysis of expression in Alzheimer's disease. Neuron 4:253-267; 1990. Haass, C.: Hung, A. Y.: Selkoe, D. J. Processing of B-amyloid precursor protein in microglia and astrocytes favors an internal localization over constitutive secretion. J. Neuroscience 11:37833793; 1991. Hajimmohammadreza, I.; Brammer. M. J.; Eagger. S.; Burns, A.; Levy, R. Platelet and erythrocyte membrane changes in Alzheimer's disease. Biochim. Biophys. Acta 1025:208-214; 1990. Harrison. P.; Wilbourn, B.; Debili, N.: Vainchenker, W.; BretonGorius, J.; Lawrie, A. S.; Masse, J.-M.; Savidge, G. F.; Cramer, E. M. Uptake of plasma fibrinogen into the a-granules of human megakaryocytes and platelets. J. Clin. Invest. 84:1320-1324: 1989. Hart, M. N.; Merz, P.; Bennett-Gray, J.; Menezes, A. H.; Goeken, J. A.: Schelper, R. L.; Wisniewski, H. M. B-Amyloid protein of Alzheimer's disease is found in cerebral and spinal cord vascular malformations. Am. J. Pathol. 132:167-172; 1988. Hawiger, J. Platelets: receptors, adhesion, secretion. Meth. Enzymol. 169:191-195: 1989. Ishii, T.; Kametani, F.; Haga, S.; Sato, M. The immunohistochemical demonstration of subsequences of the precursor of the amyloid A4 protein in senile plaques in Alzheimer's disease. J. Neuropathol. Appl. Neurobiol. 15:135-147; 1989. Joachim, C.; Games, D.; Morris, J.; Ward, P.; Frenkel, D.; Selkoe, D. Antibodies to non-beta regions of the beta-amyloid precursor protein detect a subset of senile plaques. Amer. J. Pathol. 138:373-384: 1991. Joachim, C. L.; Moil, H.; Selkoe, D. J. Amyloid ~-protein deposition in tissues other than brain in Alzheimer's disease. Nature 341:226-230; 1989. Jungi, T. W.; Spycher, M. O.; Nydegger, U. E.; Barandun, S. Platelet-leukocyte interaction: Selective binding of thrombin stimulated platelets to human monocytes, polymorphonuclear leukocytes, and related cell lines. Blood 6:629-636; 1986. Kang, J.; Lemaire, H.-G.; Unterbeck, A.; Salbaum, J. M.; Masters, C. L.; Grzeschik, K.-H.; Multhaup, G.; Beyreuther, K.; MullerHill, B. The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature 235:733-736; 1987. Kitaguchi, N.; Takahashi, Y.: Tokushima, Y.; Shiojiri, S.; Ito, H. Novel precursor of Alzheimer's disease amyloid protein shows protease inhibitory activity. Nature 331:530-532; 1988. Kisilevsky, R. In: Natrig, J. B.; Forre, D.; Husby, G.; Husebekk, A.; Skogen, B.; Sletten, K.: Westermark, P., eds. Amyloid and Amyloidosis: Vith International Symposium on Amyloidosis, Oslo, Norway, Kluwer Academic Publishers; 1990:251-253. Knops, J.; Johnson-Wood, K.; Schenk, D. B.; Sinha, S.; Lieberburg, I.; McConlogue, L. Isolation of baculovirus-derived secreted and full-length B-amyloid precursor protein. J. Biol. Chem. 266:7285-7290; 1991. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature B227:680--685; 1970. Larsen, E.; Celi, A.: Gilbert, G. E.: Furie, B. C.; Erban, J. K.; Bonfanti, R.; Wagner, D. D.; Furie, B. PADGEM protein. A receptor

433

38. 39. 40.

41.

42.

43. 44. 45. 46.

47.

48.

49.

50.

51.

52. 53.

54.

55.

that mediates the interaction of activated platelets with neutrophils and monocytes. Cell 59:305-312, 1990. Lucotte, G.; Berriche, S.; David, F. Alzheimer's mutation. Nature 351:530; 1991. Maniatis, T.; Fritsch, E. F.; Sambrook, J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989. Mann, D. M. A.; Brown, A.; Prinja, D.; Davies, C. A.; Landon, M.; Masters, C. L.; Beyreuther, K. An analysis of the morphology of senile plaques in Down's syndrome patients of different ages using immunocytochemicai and lectin histochemical methods. Appl. Neurobiol. 15:317-329; 1989. Masters, C. L.; Simms, G.; Weinman, N. A.; Multhaup, G.; McDonald, B. L.; Beyreuther, K. Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc. Natl. Acad. Sci. USA 82:4245--4249; 1985. Moenning, U.; Koenig, G.; Prior, R.; Mechler, H.; Schreiter-Gasser, U.; Masters, C. L.; Beyreuther, K. Synthesis and secretion of Alzheimer amyloid A4 precursor protein by stimulated human peripheral blood leukocytes. FEBS 277:261-266; 1990. Murrell, J.; Farlow, M.; Ghetti, B.; Benson M. D. A Mutation in the Amyloid Precursor Protein Associated with Hereditary Alzheimer's Disease. Science 254;97-99: 1991. Mustard, J. F.; Kinlough-Rathbone, R. L." Packham, M. A. Isolation of human platelets from plasma by centrifugation and washing. Meth. Enzymol. 169:3-11; 1989. Nachman, R.: Levine, R.: Jaffe, E. A. Synthesis of factor VIII antigen by cultured guinea pig megakaryocytes. J. Clin. Invest. 60:914-921; 1977. Naruse, S.: lagarashi, S.; Kobayashi, H.; Aoki, K.; Inuzuka, T.; Kaneko, K.; Shimizu, T.; lihara, K.; Kojima, T.; Miyatake, T.; Tsuji, S. Mis-sense mutation Val-lle in exon 17 of amyloid precursor protein gene in Japanese familial AIzheimer's disease. Lancet 337:879-979: 1991. Oltersdorf, T.: Firtz, L. C.; Schenk, D. B.; Lieberburg, I.; JohnsonWood, K. L.; Beattie, E. C.; Ward, P. J.; Blacher, R. W.; Dovey, H. F.: Sinha. S. The secreted form of the Alzheimer's amyloid precursor protein with the Kunitz domain is protease nexin-II. Nature 341 : 144-147:1989. OItersdorf, T.; Ward, P. J.: Henriksson, T.; Beattie, E. C.; Neve, R.: Lieberburg, I.: Fritz, L. C. The AIzheimer amyloid precursor protein. Identification of a stable intermediate in the biosynthetic/ degradative pathway. J. Biol. Chem. 265:4492-4497; 1990. Palmert, M. R.; Podlisny, M. B.: Witker, D. S.; OItersdorf, T.: Younkin, L. H.; Selkoe, D. J.; Younkin, S. G. Antisera to tan amino-terminal peptide detect the amyloid protein precursor of AIzheimer's disease and recognize senile plaques. Biochem Biophys. Res. Comm. 156:432-437; 1988. Palmert, M. R.; Podlisny, M. B.: Witker, D. S.; Oltersdorf, T.; Younkin, L. H.; Selkoe, D. J.; Younkin, S. F. The B-amyloid prorein precursor of AIzheimer's disease has soluble derivatives found in human brain and cerebrospinal fluid. Proc. Natl. Acad. Sci. USA 86, 6338-6342. Permutter, L. S.; Chui, H. C.; Saperia, D.; Athaniker, J. Microangiopathy and the colocalization of heparan sulfate proteoglycan with amyloid in senile plaques of AIzheimer's disease. Brain Research 508:13-19; 1990. Piletz, J. E.; Sarasua, M.; Whitehouse, P.; Chotani, P. lntracellular membranes are more fluid in platelets of Alzheimer's disease patients. Neurobiol. Aging. 12:401-406: 1991. Pidlisny, M. B.; Mammen, A. L.; Schlossmacher, M. G.; Palmert, M. R.; Younkin, S. G.; Selkoe, D. J. Detection of soluble forms of the ~-amyloid precursor protein in human plasma. Biochem. Biophys. Res. Commun. 167:1094-1101; 1990. Podlisny, M. B.; Tolan, D. R.; Selkoe, D. J. Homology of the amyIoid beta protein precursor in monkey and human supports a primate model for beta amyloidosis in AIzheimer's disease. Am. J. Path. 138:1423-1435; 1991. Ponte, P.; Gonzalez-DeWhitt, P.; Schilling, J.; Miller, J.; Hsu, D.; Greenberg, B.; Davis, K.; Wallace, W.; Lieberburg, I.; Fuller, F.; Cordell. B. A new A4 amyloid mRNA contains a domain homologous to serine proteinase inhibitors. Nature 331:525-527; 1988.

434 56. Ripoche, J.; Day, A. J.: Harris, J. R.; Sim, R. B. The complete amino acid sequence of human complement factor H. Biochem. J. 249:593-602; 1988. 57. Rumble, B." Retallack, R.; Hilbich, C.: Simms, G.; Multhaup, G.: Martins, R.; Hockey, A.; Montgomery, P.: Beyreuther, K.; Masters, C. L. Amyloid A4 and its precursor in Down's syndrome and Alzheimer's disease. N. Engl. J. Med. 320:1446-1452; 1989. 58. Saiki. R. K.; Gelfand, D. H.; Stoffel, S.: Scharf, S. J., Higuchi, R.; Horn, G. T.; Mullis, K. B.; Ehrlich, H. A. Primer directed enzymatic amplification of cDNA with thermostable DNA polymerase. Science 239:487-491 ; 1988. 59. Scheibel, A. B.; Duong, T.; Jacobs, R. Alzheimer'sdisease as a capillary dementia. Ann. Med. 21:103-107; 1989. 60. Schlossmacher, M. G.: Podlisny, M. B.; Ostaszewski, B. L.; Mammen, A. L.; Selkoe, D. J. Detection of circulating forms of the amyloid p-protein precursor (APP) in human plasma. Neurobiol. Aging (Abstract) 11:332; 1990. 61. Schubert, D.; Schroeder, R.; LaCorbiere, M.; Saitoh, T.; Cole, G. Amyloid B protein precursor is possibly a heparan sulfate proteoglycan core protein. Science 241:223-226; 1988. 62. Selkoe. D. J. Molecular pathology ofamyloidogenic proteins and the role of vascular amyloidosis in Alzheimer's disease. Neurobiol. Aging 10:387-395; 1989. 63. Selkoe, D. J.: Abraham, C. R.; Podlisny, M. B.; Duffy, L. K. Isolation of low-molecular-weight proteins from amyloid plaque fibers in Alzheimer's disease. J. Neurochem. 146:1820-1834; 1986. 64. Selkoe, D. J.; Bell, D.: Podlisny, M. B.; Cork, L. C.; Price, D. L. Conversation of brain amyloid proteins in aged mammals and in humans with Alzheimer's disease. Science 235:873-877; 1987. 65. Selkoe, D. J.; Podlisny, M. D.; Joachim, C. L.; Vickers, E. A.; Lee, G.; Oltersdorf, T./5-Amyloid precursor protein of Alzheimer disease occurs as 110- to 135-kilodalton membrane-associated proteins in neural and nonneural tissues. Proc. Natl. Acad. Sci. USA 85:7341-7345: 1988. 66. Silver, S. M.: McDonough, M. M.; Vilaire, G.: Bennett S. J. The in vitro synthesis of polypeptides for the membrane glycoprotein llb and ilia. Blood 69:1031-1037: 1987. 67. Sims, P. J.: Wiedmer. T.; Esmon, C. T.; Weiss, H. J.: Shattil, S. J. Assembly of the platelet prothrombinase complex is linked to vesiculation of platelet plasma membrane. J. Biol. Chem. 264:17049-17057:1989. 68. Sisodia, S. S.; Koo, E. H.: Beyreuther, K.; Unterbeck, A.; Price, D. L. Evidence that the/3-amyloid protein (/~/A4) in Alzheimer's disease is not derived by normal processing. Science 248:492-495: 1990. 69. Smith, R. P.; Higuchi. D. A.: Broze, Jr., G. J. Platelet coagulation factor Xla-inhibitor, a form ofAIzheimer amyloid precursor protein. Science 248:1126-1128; 1990. 70. Sparks D. L.: Hunsaker, II1, J. C.; Scheff, S. W.: Kryscio. R. J.: Henson, J. L.; Markesbery, W. R. Cortical senile plaques in coronary artery disease, aging and Alzheimer's disease. Neurobiol. Aging 11:601 : 1990.

S C H L O S S M A C H E R ET AL. 7 I. Swaminathan, S.; Tomkinson, B.; Kieff, E. Recombinant EpsteinBarr-virus with small RNA [EBR] genes deleted transforms lymphocytes and replicates in vitro. Proc. Natl. Acad. Sci. USA 88:1546-1550; 1991. 72. Takio, K.; Hasegawa, M.: Titani, K.; lhara, Y. Identification of~ protein precursor in newborn rat brain. Biochem. Biophys. Res. Comm. 160:1296-1301; 1989. 73. Tamaoka, A.; Kalaria, R. N.; Lieberburg, 1.; Selkoe, D. J. Identification of a stable fragment of the Alzheimer amyloid precursor containing the B-protein in brain microvessels. Proc. Natl. Acad. Sci. USA, 89:1345-1349; 1992. 74. Tanzi, R. E.; McClatchey, A. I.; Lamperti, E. D.; Villa-Komaroff, L. L.; Gusella, J. F.; Neve, R. L. Protease inhibitor domain encoded by an amyloid protein precursor mRNA associated with Alzheimer's disease. Nature 331:528-530; 1988. 75. Trip, M. D.; Manger, V. M.; van Capelle, F. J. L.; Vreeken, J. Platelet hyperreactivity and prognosis in survivors of myocardial infarction. N. Engl. J. Meal. 322:1549-1554; 1990. 76. van Nostrand, W. E.: Schmaier, A. H.; Farrow, J. S.; Cunningham, D. D. Protease nexin-ll (amyloid B-protein precursor): a platelet B-granule protein. Science 248:745-748; 1990. 77. van Nostrand, W. E.; Wagner, S. L.; Suzuki, M.; Choi, B. H.; Farrow, J. S.; Geddes, J. W.: Cotman, C. W.; Cunningham, D. D. Protease nexin-ll, a potent anti-chymotrypsin, shows identity to amyIoid B-protein precursor. Nature 341:546-549; 1989. 78. Weidemann, A.; Konig, G.; Bunke, D.; Fischer, P.; Salbaum, J. M.; Masters, C. L.; Beyreuther, K. Identification, biogenesis, and localization of precursors of Alzheimer's disease A4 amyloid protein. Cell 57:115-126; 1989. 79. Wiedmer, T.; Shattil. S. J.; Cunningham, M.; Sims, P. J. Role of calcium and calpain in complement-induced vesiculation of the platelet membrane and its exposure of the platelet factor Va receptor. Biochemistry 29:623-632; 1990. 80. Wiedmer, T." Sims, P. J. Effect of complement proteins C5b-9 on blood platelets. J. Biol. Chem. 260:8014-8019; 1985. 81. Yamaguchi, H.; Yamazaki, T.; Lemere, C. L.: Frosch, M. P.; Selkoe, D. J. Beta amyloid is deposited within the basement membrane in the early stage of amyloid angiopathy. Am J. Path. in press. 82. Yang, G. C. H.; Gallo, G. R. Protein A-gold immunoelectron microscopic study of amyloid fibrils, granular deposits and fibrillar tunnel aggregates in renal amyloidosis. Am J. Palhol. 137:1223-1231: 1990. 83. Zubenko, G. S.; Huff, F. J.; Beyer, J.: Auerbach, J.; Teply, I. Familial risk of dementia associated with a biologic subtype of AIzheimer's disease. Arch. Gen. Psychiatry 45:889-893; 1988. 84. Zubenko, G. S.; Malinakova, I.; Choinacki, B. Profliferation of internal membranes in platelets from patients with Alzheimer's disease. J. Neuropath. Exp. Neurol. 46:407-418; 1987. 85. Zubenko, G. S.; Wusylko, M.; Cohen, B. M.; Boiler, F.; Teply, I. Family study of platelet membranic fluidity in Alzheimer's disease. Science 238:539-542; 1987.