Apr 20, 1992 - amyloid protein precursor (APP) has attracted much attention ... Development Corporation of Australia (to C. L. M.). The costs ... page charges.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 267, No. 33, Issue of November 25, pp. 23950-23956, 1992 Printed in U.S. A.
Alzheimer /3A4-amyloidProtein Precursor in Immunocompetent Cells* (Received for publication, April 20, 1992)
Ursula Manning+,Gerhard KonigS,Richard B. BanatiQ, Hans MechlerS,Christian Czech+, Jochen GehrmannQ, Ursula Schreiter-GasserV, Colin L. Masters[[,and Konrad BeyreutherS From the $Center for Molecular Biology Heidelberg, University of Heidelberg, Im Neuenheimer Feld 282, W-6900 Heidelberg, Germany, the §Department of Neuromorphology, Max-Planck-Znstitute for Psychiatry, W-8033 Martinsried, Germany, the 7lCenter for Mental Health, 55, W-6700Mannheim, Germany, and the 11 Department of Pathology, University of Melbourne, ParkvilL, Victoria 3052, Australia
The mechanism of proteolytic breakdown of the BA4- in a wide variety of tissues (10, 24). It has been proposed to amyloid protein precursor (APP) has attracted much play a role in cell-cell interactions and cellular growth (6-8). attention because of its relevance for Alzheimer’s dis- APP is secreted by several cultivated cell lines, and secreted ease. Apart from the pathological role of APP in the APP isoforms have also been found in plasma and cerebroamyloidogenesis, many efforts havebeen made to iden- spinal fluid, as well as other extracellular compartments (9tify the functional significance of this widely expressed 12). It was therefore suggested that APP may mediate addiprotein in various biological processes. Employing bio- tional cell-adhesive interactions in diverse biological procchemical techniques, we demonstrate that APP is in- esses, which includes inflammation, immune response, and volved in the initiation of the immune response. Upon stimulation, it is expressed by the major functional regeneration. Previously we were able to show induction of types of T-lymphocytes, i.e. CD4+ and CDS+ cells. As APP secretion in stimulated leukocytes suggesting a role for was demonstrated for the CD4+ lymphoid cell lineHS, APP in the construction of the immunological network (13). APP is predominantlysecreted. The remaining COOH- The soluble form of APP, secreted by platelets, is identical terminal fragments generated upon secretion were with a serine protease inhibitor, protein nexin I1 (PN 11),the highly unstable. Of the APP produced by immunocom- natural inhibitor of the blood coagulation factor XIa (14-18). petent cells, considerable amounts were shown to be Therefore, APP may also participate in wound healing/tissue leukocyte-derived APP (L-APP). In addition, we were repair of the vasculature. able to identify the KPI-containing L-APP isoform, LAPP constitutes a family of different isoforms that are APP733, as the major expressed L-APP isoform in generated by alternative splicing. The major primary transimmunocompetent cells, including rat microglial cells lation products consist of 695, 751, and 770 amino acid resiand astrocytes. The L-APP expression pattern of these dues (APP695, APP751, and APP770), respectively. The cells showed high similarity. These findings seem to be amino acid sequences of these APP isoforms show the charindicative of an important function of APP within the acteristicfeatures of typical membrane glycoproteins. Seimmune system. Therefore, APP may be involved in creted forms of APP are generated by proteolytic cleavage various immunopathogenic conditions of the periphery within the amyloidogenic region(19,20). Theyall contain the and inthe central nervous system. complete and partially membrane-inserted PA4 sequence (2124). APP751 and APP770, but not APP695, contain a protease-inhibitor domain, giving rise to the serine proteaseinhibitor (KPI) function of APP (14, 16, 22, 23). The pA4-amyloid protein precursor (APP)’ is thought to In additionto theubiquitously expressed APP isoforms, we be processed aberrantly to thePA4 protein, the major proteinidentified recently a leukocyte-derived splice form, which was aceous component of amyloid deposits, which are the key termed L-APP after leukocyte-derived APP (25). L-APPisofeatures and major diagnostic lesions in brains of individuals forms are APP transcripts lacking exon 15, which leads to an with Alzheimer’s disease (1-4). The deposits of abnormal PA4exclusion of 18 amino acids, generating isoforms of APP695, amyloid proteins are observed in intracellular and extracel-751, and -770 with 677, 733, and 752 residues, respectively. lular locations of the brain and may lead to neuronal dysSince we demonstrated aregulated alternative splicing of exon function and final neuronal death (3-5). APP is a multifunctional glycoprotein that is synthesized 15 in leukocytes in response to an immunogenic stimulus, we initially proposed that L-APPsplay a role in initial events of * This work was supported by grants SFB 317 and 258 from the cell targeting (25). The identification of L-APP mRNA Deutsche Forschungsgemeinschaft, the Bundesminister fur For- expression inrat brain macrophages (activated microglial schung und Technologie of Germany, the Metropolitan Life Foun- cells)’ supports this hypothesis, because the appearance of dation, the Thyssen Stiftung,the Fonds der Chemischen Industrie of activated microglial cells is known to be associated with Germany, the Forschungsschwerpunkt Baden Wurttemberg (to K. various neuropathological states including neurodegenerative B.),the National Health and Medical Research Council of Australia, the Victorian Health Promotion Foundation, and the Aluminium and inflammatory diseases in human brain (25-29). We have examined the expression of APP/L-APP during Development Corporation of Australia (to C. L. M.). The costs of publication of this article were defrayed in part by the payment of immune mediated responses in freshly isolated leukocytes page charges. This article must therefore be hereby marked “adver- since these cells offer several advantages over other cells for tisement’’ in accordance with 18U.S.C. Section 1734 solely to indicate analysis of APP biogenesis and metabolism. To analyze the this fact. The abbreviations used are: APP, PA4-amyloid protein precursor; lymphocyte-derived APP metabolism, we have used the PMBL, peripheral mononuclear blood leukocyte; PHA, phytohemag- lymphoid cell line H9, which exhibits high APP expression. glutinin; PMSF, phenylmethanesulfonyl fluoride; KPI, serine proteinase inhibitor; PAGE, polyacrylamide gel electrophoresis; bp, base pair(s).
R. Banati, J. Gehrmann, C. Czech,U. Manning, G. Konig,K. Beyreuther, and G. W. Kreutzberg, manuscript submitted.
23950
23951
APP in Leukocytes We also examined APP expression in different cells of the nervous system and compared their APP-specific biosynthetic behavior to that of T-lymphocytes. From these studies we conclude that APP is associated with defense mechanisms in the central nervous system involving the intrinsic immune system. EXPERIMENTALPROCEDURES
Cell Lines The lymphoid cell line H9 (31) (kindly provided by P. Gehrmann, University of Dusseldorf) was cultured in medium (RPMI 1640) containing15%fetal calf serum and glutamine. HeLa cells were maintained in Dulbecco'smodifiedEagle's medium supplemented with 15% fetal calf serum (GIBCO/Bethesda Research Laboratories (BRL)) andglutamine. Cell Preparation and Culture Conditions Leukocytes-Human peripheral mononuclear blood leukocytes (PMBLs) from venous blood of human volunteers (Zentralinstitut fur seelische Gesundheit, Mannheim) or from buffy coats of normal donors (Blood center, Heidelberg) were prepared by Ficoll-Hypaque gradient density centrifugation (32). CD4+ and CD8' lymphocytes were isolated via magnetizable polymer beads coupled to monoclonal antibodies against the specific surface antigens (Dynabeads, Dynal) (54). Isolated leukocytes were cultured at a density of 2 X lo6 cells/ ml in Dulbecco's modifiedEagle's medium (GIBCO/BRL) containing 1 g/liter glucose, penicillin (50 units/ml), streptomycin (40 pglml), and 10% (v/v) fetal calf serum. Activation of resting PMBLs was induced by culturing cells in medium supplemented with 10 pg/ml phytohemagglutinin (PHA, Sigma). CD4+ and CD8+ lymphocytes were activated with conditioned medium of PHA-stimulated PMBLs in combination with PHA. Medium was not exchanged during stimulation. Microglia and Astrocytes-Isolation of microglial cells and astrocyte-enriched cultures from rat brain and culture conditions was as previously described (33).
buffer A (20 mM Tris, pH 7.5, 150 mM NaCI, 0,4% Nonidet P-40, 0.4% Triton X-100, 2 mM PMSF), washing buffer B and finally with TSA solution (20 mM Tris, pH 8.0, 150 mM NaC1). The immunoprecipitates were fractionated by SDS-PAGE (34). Gels were soaked in enhancer solution (En3Hance;Du Pont-New England Nuclear), dried, and exposed to Kodak X-Omat AR film at -70 "C. Secreted, nonlabeled APP was analyzed by immunoprecipitation and subsequent immunoblotting. The precipitation was done under nondetergent conditions. Nonbound proteins wereremovedfrom Sepharose beads by washing three times with TSA solution. Immunoblotting Immunoblotting was performed using chloroform/methanol-precipitated cell lysates of H9 cells (35) or immunoprecipitates of supernatants.The samples were fractionated by SDS-PAGE (34) and transferred to nitrocellulose (36). The nitrocellulose sheet was soaked in phosphate-buffered saline containing 1%bovine serum albumin for 1 h at room temperature and thenincubated a t room temperature for 2 h with affinity-purified anti-CT IgG (0.3 pg/ml in phosphatebuffered saline, 1%bovine serum albumin) or monoclonal antibody 22Cll. The nitrocellulose sheet was washed three times with Tris-buffered saline (10 mM Tris, pH 8.0, 150 mM NaC1,0,05% Tween 20). Specific binding was visualized with the Protoblot alkaline phosphatase system (Promega, Heidelberg). Isolation of RNA and SI Nuclease Protection Analysis Total RNA was prepared as described (37). A1075-bp APP cDNA fragment was generated with mutated primers by polymerase chain reaction from position +690 (with a SphI restriction site) toposition +1761 (with an EcoRI restriction site) of the APP695 cDNA. This polymerase chain reaction fragment was digested with SphIand EcoRI, which gives rise to a 1045-bp APP695 cDNA fragment. This fragment was then subcloned into the phage M13 mp19, resulting in recombinant phage MJu695 (38). Generation of the S1probe and S1 nuclease protection analysis was performed as previously described (25). Cell Transfection
Biosynthetic Labeling After removal of the culture medium, 1 X lo7 cells were radioactively labeled with 120 pCi of [35S]methionine(Amersham) in 1.5 ml of methionine-free Dulbecco's modified Eagle's medium for 4 h. For pulse-chase experiments, cells were pulsed for 8 min with 150 pCi of ["Slmethionine and chased for different time intervals with an excess of unlabeled methionine (1 mM). The conditioned medium was then cleared by centrifugation and stored a t -20 "C. The cells were washed once with phosphate-buffered saline. For lysis, cells ( 5 X lo6 cells) were resuspended in 0.2-ml lysis buffer (50 mM Tris, pH 7.5, 150mM NaCl, 2 mM EDTA, 1% Nonidet P-40, 1%Triton X-100, 2 mM phenylmethanesulfonyl fluoride (PMSF), 10 pg/ml aprotinin, 10 pg/ ml leupeptin, and 1.6 mg/ml iodoacetamide) and incubated for 30 min on ice. Cell lysates were centrifuged at 10,000 X g for 5 min, and the supernatantswere stored at -20 "C until further use. The extraction pellet was discarded. Immunoprecipitation For immunoprecipitation of [35S]methionine-labeledAPP, 100 p1 of cell lysate or 750 pl of conditioned medium were used. The cell lysate was diluted to a concentration of 1:4 with ice-cooled washing buffer B (20 mM Tris, pH 7.5, 500 mM NaCl, 0,5% Nonidet P-40, 2 mM PMSF), and the conditioned medium was supplemented with 75pl medium buffer (1M Tris, pH 8.0, 100 mM EDTA, 10% Nonidet P40, 100 mM PMSF). Lysate and medium were preincubated for 2 h a t room temperature with 10 p1 of preimmune serum and 3 mgof protein A-Sepharose (Pharmacia, Uppsala, Sweden). The insoluble complexes were centrifuged down and discarded. The supernatants were incubated for 1 h at room temperature with 5 pl of undiluted anti-FdAPP or anti-CT antiserum. Anti-FdAPP is polyclonal rabbit antiserum raised against purified Escherichia coli FdAPP fusion protein consisting of APP695 and theFd fragment of the murine IgM immunoglobulin heavy chain (10). The polyclonal anti-CT antiserum was raised against asynthetic peptide, which corresponds to the COOH-terminal43 residues of APP. After addition of 2 mg of protein A-Sepharose, the mixture was incubated for 30 min at room temperature. Nonbound proteins were removed from Sepharose beads by sequential washing with washing
For transient expression of L-APP, monolayers of HeLa cells were transfected with 20 pg of CMV expression vector, which contains LAPP cDNA inserts (10,25)using Lipofectin Reagent (BRL) according to the manufacturer's instructions.L-APPinsert cDNA used for transfection was obtained by polymerase chain reaction of reverse transcribed (oligo(dT)) mRNA from human peripheral lymphocytes (25). For analysis of L-APP expression, cells were radiolabeled for 3 h with 120 pCi of [35S]methionine48 h after transfection and subjected to immunoprecipitation analysis as described above. RESULTS
APP Secretion by Leukocytes-PHA-stimulated human PMBLs secrete APP. Previous experiments have shown that, in contrast to resting PMBLs, stimulation of PMBLs with mitogens such as PHA, concanavalin A, and pokeweed mitogen induces APP secretion (13). The generated secretory APP isoforms quantitatively outrankthe membrane-bound species. Differences in the amount of secretory APPs detected by comparison of APP secretion of stimulatedPMBLs from patients with Alzheimer's disease and healthy control donors were not correlated with a clinical diagnosis of Alzheimer's disease (data not shown). Expression of APP by CD4' and CD8' Lymphocytes-Since PHA stimulates mainly T-lymphocytes, we have originally suggested that activated T-lymphocytes arethe origin of secreted APP in supernatants of PMBLs (13). However, the types of T-cells producing APP remain unknown. There are at least two functionally distinct populations of T-lymphocytes, the cytotoxic T-cells (Tc) and the helper T-cells (Th), which express the CD8 and CD4 surface antigens, respectively. Therefore, we have examined whether thereexistsa difference in APP expression in these two functional subsets of T lymphocytes. Highly enriched populations ofCD4' and CD8+ cells were prepared by positive
23952
APP in Leukocytes cellmedium lysate
FIG. 1. APPbiosynthesis of PHA-stimulated CD4+ and CDS' lymphocytes. Freshlyisolated CD4' and CD8' cells were cultivated for 3 days in presence of PHA in combination withconditioned medium of PHA-activated PMBLs. Biosyntheticlabeling with [:"S]methionine was performed after 3 days of cultivation. Cells and conditioned medium were subjected to immunoprecipitationwith anti-FdAPP. The precipitates were analyzed by SDS-PAGE (8%) and autoradiography. Specific bands corresponding to APPisoforms are indicated.
selectionusing magnetizable polymer beadsconjugated to anti-CD8 or anti-CD4 monoclonal antibody (55). Stimulation was performed usingsupernatants of PHA-treated PMBLs in combinationwith PHA for 72h. By immunoprecipitation, significant quantities of APP could be detected intracellularly as well as in the conditioned media of activated CD4' and CD8' cells (Fig. 1).Using polyclonal anti-FdAPP sera, immunoprecipitation of predominantly intracellular APP isoforms of approximately 140 and 105 kDa and small amounts of APP in the range of 125 and 135 kDa were observed. The bands may represent biosynthetic full length as well as COOH-terminal-truncated,mature APPisoforms (10). The predominant APP species secreted by CD4' and CD8' lymphocytes have a molecular mass of approximately125 kDa. Upon longer exposure, a more complex band pattern due to L-APP isoforms became apparent (see below). We suggest that the nearly identical expression of APP by CD4' and CD8' cells may indicate a participation of APP in fundamental processes of the immunesystem, rather thana more specialized function of APP for the two subsets of T-cells. Analysis of A P P Biosynthesis in H9 Cells-For a further analysis of APP biosynthesis in T-lymphocytes we screened several human T-lymphoid cell lines for a high level of APP expression. The T-cell line H9 was found to constitutively express high levels of APP (Fig. 2). H9 was originally derived from a HUT 78 cell line, ahuman cutaneousT-cell lymphoma (31). The human T-cell lines CEM, Molt-4, and HPB-ALL do not express detectable amounts of APP? To analyze the metabolic behavior of endogenous APP expression in H9cells, we performed an immunoprecipitation after labeling with [35S]methionine. As shown in Fig. 2b, the H9 lymphoid cells expressed predominantly APP isoforms containing the COOH terminus (molecular mass range of 105-140 kDa). An additional holo-APP of approximately 135 kDa was also detected (Fig. 2b). COOH-terminally truncated APP, not precipitated by the anti CT-antiserum,was identified as a band with the molecular mass of approximately 125
'H. Mechler and U. Manning, unpublished results.
a. b. C. FIG. 2. APP biosynthesisof H9 cells. Autoradiography of APP isoforms precipitated from H9 cells. APP of 3sS-radiolabeled cellsand media was immunopreciptated with anti-Fd APP antiserum (a, c) or anti-CT antiserum ( b ) ,respectively. The precipitates were subjected t o SDS-PAGE (8%)and autoradiography.
kDa (Fig. 2a). This band has the same electrophoretic mobility as the major APP isoforms detected in the conditioned medium of H9 cells and therefore corresponds to secretory APP (Fig. 2c). An identical APP pattern was found for freshly isolated and stimulatedCD4' and CD8' cells (Fig. 1). To study APP maturation, a pulse-chase experiment was performed. H9 cells were labeled with [35S]methionine for 8 min and then "chased" with 10 mM methionine for varying periods of time. APP was immunoprecipitated from cell lysates and conditioned medium and analyzed by SDS-PAGE. Initially cell-associated immunoreactivity was detected as a 105-kDa band (Fig. 3, cell lysate). The intensity of this 105kDa APP band gradually decreased during the chase period. New APP isoforms, migrating at 135 and 140 kDa appeared within the first 10 min of chase in comparable amounts. This indicates a posttranslational modification, as previously described by Weidemann et al. (1989) (10) for other cell lines. The amounts of these higher APP isoforms remained nearly constant between 20 and 70 min of chase. Secreted APP asa double band at a molecular mass range of approximately 120125 kDa was already detected within 20 min of chase (Fig. 3, medium), suggesting a very low half-life time of matured cellassociated APP. During longer chase times, the upper APP band at 125 kDa became thepredominant secreted APP isoform. Identification of L-APP-We previously identified a leukocyte-derived APP mRNA (L-APP) inT-cells, macrophages, and microglial cells (25). To confirm the identity of APP mRNAs in lymphocytes, total RNA from lymphocytes was isolated and analyzed by S1 nuclease protection analysis. The S1 probe was constructed in such a way that we were able to distinguish APP695 mRNA from L-APP677 mRNA and the KPI-containing transcripts APP751/770 from APP733/752 (Fig. 4a). As shown in Fig. 4, in activated lymphocytes the predominant APP transcripts were APP751/770 mRNA. The L-APP733/752 transcripts were detected with relative amounts of approximately 30% of total APP mRNA (Fig. 4b). Only minor amounts of the KPI-lacking transcripts APP695 and L-APP677 were visible (Fig. 46). To identify and further characterize the corresponding LAPP protein isoforms, constructs encompassing L-APP cDNA sequences (L-APP677, L-APP733, and L-APP752)
9
23953
APP in Leukocytes lysate
cell 0
10
20
30
40
506070
90
kb
0
10
20
3 0 4 0 5 0 6 70 0 9 0
chase time [min]
a=
FIG. 3. Kinetic analysis of APP maturation. H9 cells were pulse-labeled for 8 min with [:%]methionine followed by addition of an excess of unlabeled methionine (10 mM). At various times of chase (0-90 min) aliquotswere removed, and cells and conditioned medium were subjected to immunoprecipitation, SDS-PAGE(8%),and autoradiography. Immunoprecipitations were performed with anti-FdAPP anitserum (precipitated proteins from cell lysate of the time point 60min were degradated during preparation). SecretedAPP isoforms are indicated by arrows. The upper band corresponds to secretory APP751, the lower band corresponds to L-APP733.
A PL S i probe: 5 . 1 6
PL I
9
1101
11
J
protected fragments: APP 695
I 12 1 1 3 1
14
hybridization, S1 digestion
L-APP 677
n 5 a 3'
bases 1045
978
APP 7701751 L-APP 752l733
B
APP 695 L-APP 677
FIG.4. a, SI probe MJu695. The S1 probe MJu695 consists of a 1045-bp APP695 cDNA fragment comprising part of exon 6, exons 9-15, and partof exon 16 (denoted asboxes). PL shows the polylinker, which is derived from the phage M13 mp19. After hybridization and fragments of 1045 bases for subsequent S1 nucleasedigestion, APP695, 978 bases for L-APP677, 886 bases for APP7701751, and 819 bases for L-APP752/733, respectively, are protected. b, alternative splicing of exons 7, 8, and 15 in human T-lymphocytes. Total RNA (20 pg) from healthy human T-lymphocytes, stimulated for 5 days with phytohemagglutinin, was hybridized overnight to uniformly labeled probe MJu695. Hybrids were subjected to SI nuclease digestion and analyzed on a sequencing gel. The autoradiograph of a 48-h exposure timeis shown. Protected fragmentsfor APP695, L-APP677, APP7701751, and L-APP7521733 mRNAs respectively, are visible.
were cloned into an eukaryotic expression vector and transfected into HeLacells (10). Upon 48 h of transient expression, cells were metabolically labeled with [35S]methionineand cell lysates of the different L-APP transfectants were prepared (Fig. 5). Then L-APPisoforms were immunoprecipitated from the conditioned media and thecell lysates. Full-length (transmembrane) APP, precipitated from nontransfectedHeLa cells, migrated with apparent molecular masses of 105 and
140 kDa, whereas secretoryisoforms were detected at approximately125 kDa (Fig. 5, HeLa). These predominantly expressed endogenous APP isoforms corresponded to APP751 isoforms (10). HeLa cells transfected with L-APP cDNA expression constructscontained additional cell-associated APP isoforms, which corresponds to intermediateorfinal products of L-APP biogenesis (Fig. 5). Substantial quantities of the different L-APP isoforms were secreted into the medium. Transmembrane L-APP677 migratesbelow endogenous immature APP751 (105 kDa) as an intense band at approximately 100 kDa and itssecreted counterpart at approximately 90 kDa (Fig. 5). The expected higher molecular mass of mature L-APP isoforms can be detected in thecell lysates and supernatants of the L-APP733 (135 and 120 kDa) and L-APP752 (138 and 123 kDa) transfectants (Fig. 5). H9 cells express L-APP733,which was demonstrated in the analysis of its cell lysate,where an additionalband was present when compared to HeLa cells (Fig. 5). This was also confirmed by RNA analysis (data not shown). Since H9 cells expressed and secreted predominantly APP751, and due to the small (5 kDa) difference in molecular mass between mature APP and L-APP, secreted L-APP isoforms were difficult to detect. However, in pulse-chase experiments, secreted L-APP733 was clearly distinguishable from secreted APP751 (Fig. 3). Secreted L-APP733 corresponded to the lower weak band of a double band detected by autoradiography of APPspecific immunoprecipitates of the supernatants. Effect of PHA on APP Secretion-PHA induced an alteration of the ratio of cellular APP isoforms and enhanced APP secretion. First, a marked decrease of high molecular weight isoforms was observed. The cell-associated 135- and 140-kDa APP-isoforms,corresponding tomaturetransmembrane APP751 andL-APP733 (Fig. 6,a and b ) , as well as theCOOHterminally truncated isoform (Fig. 6a) became less prominent following PHA treatmentwhen compared to untreatedcontrol cells. In contrast, therewas no apparent PHA-induced difference in the 105-kDa transmembrane immature APP-isoform between induced and noninduced cells (Fig. 6, a and b ) . Concomitant with the decrease of the cell-associated 135- and 140-kDa APP isoforms upon PHA treatment, an increase of the amount of secretory APP was observed (Fig. 6c). These observations suggested that PHA is not only a potent stimulator of APP secretion by peripheral blood lymphocytes (13), but also affects the regulation of APP secretion of the Tlymphoid cell line H 9. Since APP secretion is due to COOH-terminal proteolysis of transmembrane APP, we have conducted Western blot analysis of noninduced (C) and PHA-induced H9 cells (PHA) in order to detect remaining COOH-terminal fragments described by Esch etal. (19), Sisodia et al. (20), and recently by
APP in Leukocytes
23954
medium
cell lysate
FIG.5. Identification of molecular forms of L-APP by transfection experiments with HeLa cells. Bio-
~9
HeLa
HeLa transf. 677
synthetic labeling of L-APP forms expressed by HeLa-cell transfectants. Monolayers of HeLa cells were transfected with cDNA (HeLa trans/.) of LAPP677 (677), L-APP 733 (733), andLAPP 752 (752).Detergent extracts of HeLA transfectants and culture supernatants were then analyzed by immunoprecipitation with anti-CT serum (cell lysate) or anti-FdAPP serum (medium). For comparison, the analysis was concomitantlydone with cell lysates and supernatants of normal, nontransfected HeLa cells (HeLa) and H9 cells (H9). Mature intracellular and extracellular LAPP isoforms are indicated with arrows.
733
HeLa
I
752
kD 200
HeLa transf. 677 752 733
r
e L - A P P 752 L-APP 733
F”
+-L-APP677
e::;;; :::
100 92
CL-APP 677
68
~
medium
cell lysate
c
PnA
c
PnA
cn
C
PHA
PHA
C
3h 5h 7h 7 kd
. ”“ . I
c
a.
b.
-
-
43
-
25
-
18
-
14
I
C.
FIG.6. Effect of PHA on APP secretion. H9 cells were metabolically labeled with [%]methionine without (C)and in the presence of PHA ( P H A ) .For detection of APP isoforms, detergent extracts of cells and conditioned media were subjected to immunoprecipitation with anti-CT( b ) and anti-FdAPP(a, c) and analyzed by SDS-PAGE (8%)followed by fluorography.
Golde et al. (39). We used a monospecific anti-CT antibody. FIG. 7. Effect of PHA on stability of APP-specific COOHAs shown in Fig. 7, APP-specific COOH-terminal fragments fragments. H9 cell extracts were prepared at various migrating at approximately 14 kDa slowly decreased during terminal PHA-induction times indicateda t the topof each lane ( C ,nontreated the stimulation process. These experimentsmay indicate that H9 cells). The analysis of chloroform/methanol-precipitatedproteins PHA co-stimulates a degradation of COOH-terminal APP was performed by SDS-PAGE using a lowergel containing 15% polyacrylamide, followed by Western blot analysis using the monofragments. Preliminary dataletusto conclude that APP-specific specific anti-CT antibody. The APP-specific COOH-terminal fragments are indicated by an arrow. COOH-terminal fragments of activated human PMBLs are indeed highly unstable, since it was not possible to identify COOH-terminal APP fragments in these cells in our assay cells, astrocytes, and microglia, indicating possible differences in the ratio of APP/L-APP expression or in its biogenesis. system: APP in Immunocompetent Cells of the Brain-In order to Unstimulated H9 cells showed comparable amounts of iminvestigate a possible involvement of APP and L-APP in mature and maturetransmembrane APP isoforms,which neuroimmune reactions, activated microglial cells and astro- were visible as bands of approximately 105, 135,and 140 kDa. cytes of rat brain were isolated and tested for the presence of However, the majority of detectable intracellular APP in especially, in microglialcells reAPP and L-APP by immunoprecipitation following biosyn- culturedastrocytesand, mained immature, which can be seen as the broad 105-kDa thetic labeling with [35S]methionine.As previously shown by others (40, 41),both activated microglial cells and astrocytes band. In addition, it seemed that L-APP733 expression in express APP in high amounts. These cell types also expressed these cells contributed to a greater proportion of total APP the band pattern that resembled that of lymphocytes, indi- expression, when compared to H9 cells (see also Fig. 5). To detect secretory APP released by microglial cells and cating synthesis of L-APP733 (Fig. 8). However, differences were detected in the ratio of the distinct bands between H9 astrocytes, conditioned medium was investigated. Since the polyclonal anti-FdAPP did not precipitatesoluble isoformsof rat APP under the high detergent concentrations necessary U. Monning, unpublished results.
APP in Leukocytes
23955
We observed that APP was very unstable under these conditions. Therefore, only a minor fraction of APP secreted by microglial cells was detected by Western blot analysis. The major fraction may have been degraded. DISCUSSION
A series of experimental evidence clearly indicated that soluble APP is an importantcomponent of different biological processes such as inflammation and immune response (13,18, 25). The fact that APP as an inducible molecule is secreted by lymphocytes directly upon stimulation with mitogenic lectins, such as PHA, also suggests a role of APP during immunopathogenic processes. CD4' as well as CD8' cells were shown to secrete APP at comparable levels upon activation, which also may indicate that APPparticipates in a generalized and fundamental process of T-cell-dependent immune response. Kinetic experiments, with regard to APP maturation by lymphocytes on the lymphoid T-cell line H9 (Fig. 3) indicated that APPof lymphocytes primarily has afunction as secretory APP. This is evidenced by the fact that secreted APP could FIG.8. Comparative analysis of APP biosynthesis in lym- be detected within 20 min of chase. Consequently, membranephocytes, microglial cells, and astrocytes. Cell extracts of [%I spanning APP must have a very low half-life. APP secretion radiolabeled rat microglial cells and rat astrocyte-enriched cultures, was reported to be associated with the appearance of APPas well as of human lymphocytes (H9cells) were analyzed on transspecific COOH-terminal fragments, which result from protemembrane APP forms by immunoprecipitation with anti-CT antiserum followed by SDS-PAGE and fluorography. Molecular forms of olytic cleavage within the PA4 domain of APP (19, 20). We therefore tested the hypothesis that increased APP secretion L-APP733 are indicated by arrows. is associated with the accumulation of COOH-terminal fragments. We treated the T-lymphoid cell line H9 with the mitogenic lectin PHA and observed a drastic increase in APP secretion that may be due to increased APP-secretase activity. However, we have not detected a concomitant increase of detectable COOH-terminal fragments. On the contrary, we observed a decrease in the amount of detectable COOHterminal fragments, which may suggest that APPsecretion is physiologically linked to a mechanism responsible for simultaneous degradation of the COOH-terminal fragments that are generated by APP secretase. This raises the question whether such a secretase-linked degradation pathway also exists in neuronal cells and opens the possibility that differences instability of COOH-terminal APP fragments may have fundamental consequences for the pathogenesis of Alzheimer's disease. Preliminary experimental data performed by us have indicated that indeed COOH-terminal fragments of APP are more stablein neurons than in cells of the periphery: A similar observation of different levels of COOHFIG.9. Identification of secetory APP forms of microglial terminal APP derivatives was made by Estus et al. (1992) cells and astrocytes. Conditioned medium of microglial cells and (43), who compared homogenates of several brain regions astrocyte-enrichedcultures was subjected to immunoprecipitation with that of peripheral tissues like kidney and liver. Differwith the antiFdAPP serum. The immunopreciptates were analyzed ences in stability maybe the cause of restriction ofPA4by immunoblotting. Detection of APP was performed with the mono- amyloid depositions to thebrain of patients with Alzheimer's clonal antibody 22Cll (1:10000). disease. Apart from lymphocytes, monocytes, macrophages, and for radioactive immunoprecipitations followedby autoradi- platelets are also able to secrete APP upon stimulation (15, ography, secretory APP was identified by Western blot analy- 17, 18, 54). The detection of APP biosynthesis in immunosis with the monoclonal antibody 22Cll. Broad immunoreac- competent cells of the central nervous system, which are tive bands in the molecular mass range of 120-125 kDa were mainly microglial cells and possibly astrocytes, opens the identified (Fig. 9). These bandsshowed the same electropho- possibility that APP is involved in initial cell activation and retic mobility as thesecreted APP of H9 cells (or circulating cell-targeting events that are comparable to that of the imAPP of blood). In contrast to secreted APP of H9 cells, mune system of the periphery. The presence of activated additional bands of approximately 100 kDa were detected in microglial cells and astrocytes was described at sites of central conditioned media of microglial cells and astrocytes. They nervous system damage (44,45). Thebrain has been considappeared to be secreted forms of APP695 or proteolytic frag- ered for a long time as an immunologically privileged site ments of the KPI forms, respectively. because of the lack of a true lymphatic system andthe Since microglial cells produce high amounts of free radical- existence of several barriers that isolate it from the periphery superoxide anions in vitro, the pHof the conditioned medium of microglial cells was shifted toward acidic conditions (42). U. Monning and M. Dichgans, unpublished results.
APP in Leukocytes
23956
(30). However, it is known that astrocytes and microglial cells mediate immune reactions in the brain (for review see Ref. 46). The molecular and cellular characteristics of the immune and neuroimmune responses suggest that many elements of each system share common features, like expression of major histocompatibility complex 11, antigen presentation, and interleukin release. The identification of the novel leukocytederived isoform L-APP (25) in nearly all cells with immunerelated functions substantiatesthis hypothesis. Wecould demonstrate that activated lymphocytes as well as cultured microglial cells and astrocytes express significant amounts of the KPI-containing L-APP isoform, L-APP733. In view of the potential importance of adhesive interactions during inflammation and immune response, much work has focused on the identification of adhesion molecules produced by lymphocytes. Adhesive moleculesplay a centralrole within the immune system. They are involved in tissue targeting, activation, and cell recognition events of lymphocytes (for review see Ref. 47). The function of APP is not yet known in detail, but growing evidence exists that APP may mediate cell interactions with cell surface or soluble glycoproteins (6, 7,8,48-53). Although secreted APP and transmembrane APP belong to the same family, it could beenvisaged that they areinvolved in different cell-cell interaction activities. It seems an intriguing hypothesis that transmembrane APP could function directly as a cell-adhesion molecule (48,49, 51), while secreted APP may be involved in indirect cell interactions by generating intracellular signals important for regulation of cell interactions via a yet undefined cell-surface receptor. Since APP expression is not restricted to lymphocytes but expressed ubiquitously, we postulate that APP may beinvolved in initial, perhaps weak adhesion to target cells and, additionally, in the mediation of signals activating specific and strong cell-cell interactions. The weak adhesion maybe due to adhesive interactions of transmembrane APP to extracellular matrix components such as collagen, laminin, and proteoglycans followed by binding of secretory APP to a putative receptor on the target cells (52, 53). This APP-receptor interaction may trigger a cell-specific adhesion cascade as exemplified by integrins, selectins, and members of the immunoglobulin superfamily, thereby generating a specific intracellular activation signal (47). We postulate that L-APPplays an important role in cellular interactions associated with the rapid transition of resting leukocytes, microglia, and astrocytes to active, immunocompetent cells in response to disease, infection, or injury. APP could thus be involved in themodulation of immune responses in theperiphery and in the centralnervous system. Acknowledgments-We thank Xenia Manzel and Julia Albrecht for technical assistance. REFERENCES 1. Masters, C. L., Simms, G., Weinman, N. A,, Multhaup, G., McDonald, B. L., and Beyreuther, K. (1985) Proc. Natl. Acad. Sci. U. S. A. 8 2 , 42454249 2. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., and Stadlan, E. M. (1984) Neurology 34,939-944 3. Muller-Hill, B., and Beyreuther, K. (1989) Annu. Reu. Bioehem. 5 8 , 287307 4. Selkoe, D. (1991) Neuron 6,487-498 5. Masters, C. L., and Beyreuther, K. (1991) Brain Pathol. 1, 226-227 6. Shivers, B. D., Hilbich, C., Multhaup, G., Salbaum, J. M., Beyreuther, K., and Seeburg, P. (1988) EMBO J. 7,1365-1370 7. Saitoh, T., Sundsmo,M., Roch, J.-M., Kimura, N., Cole, G., Schubert, D., Oltersdorf, T., and Schenk,D. B. (1989) Cell 68,615-622 8. Schubert, D., Jin, L. W., Saitoh, T., and Cole, G. (1989) Neuron 3, 689694
9. Rumble, B., Retallack, R., Hilbich, C., Simms, G., Multhaup, G., Martins, R., Hockey, A,, Montgomery, P., Beyreuther, K., and Masters, C.L. (1989) N. Engl. J. Med. 3 2 0 , 1446-1452 10. Weidemann, A., Konig, G., Bunke, D., Fischer, P., Salbaum, J.-M., Masters, C. L., and Beyreuther, K. (1989) Cell 57,115-126 11. Palmert, M. R., Podlisny, M. B., Witker, D. S., Oltersdorf, T., Younkin, L. H., Selkoe, D. J., Younkin, S. G. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,6338-6342 12. Podlisny, M. B., Mammen, A.L., Schlossmacher, M. G., Palmert, M. R., Younkin. S. G.. and Selkoe. D. J. (1990) Biochern. Bioohvs. Res. Cornmun.
., G., Prior, R., Mechler, H., Schreiter-Gasser, U., reuther, K. (1990) FEBS Lett. 277,261-266 14. Oltersdorf; T., Fritz, L. C., Schenk, D. B., Lieberburg, I., Johnson-Wood, H. F.. and K. L.. Beattie. E. C.. Ward. P. J.. Blacher. R.W..Dovev. Sjnha, S. (1989) Nature, 341,~1441147 ' 15. Smlth, R. P., Higuchi, D. A,, and Broze, G. J. (1990) Science 2 4 8 , 11261128 16. Van Nostrand, W. E., Wagner, S. I., Suzuki, M., Choi, B. H., Farrow, J. S., Geddes, J. W., Cotman, C. W., and Cunningham, D. D.(1989) Nature 341,546-549 17. Van Nostrand, W. E., Schmaier, A. H., Farrow, J. S., and Cunningham, D. D. (1990) Science 2 4 8 , 745-748 18. 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., and Masters, C. L (1990) J. Biol. Chem. 266,15977-5983 19. Esch, F. S., Keim, P. S., Beattie, E. C., Blacher, R. W., Culwell, A. R., Oltersdorf, T., McClure, D., and Ward, P. J. (1990) Science 2 4 8 , 11221124 20. Sisodia, S. S., Koo, E. H., Beyreuther, K., Unterbeck, A., and Price, D. L. (1990) Science 248,492-495 21. Kang, J., Lemaire, H. G., Unterbeck, A., Salbaum, J. M., Masters, C. L., Grzeschik, K. H., Multhaup, G., Beyreuther, K.,and Muller-Hill, B. (1987) Nature 326,733-736 22. Ponte, P., Gonzalez-DeWhitt, P., Schillin , J , Miller, J., Hsu, D., Greenber , B., Davis, K., Wallace, W., Liebegurg, I., Fuller, F., and Cordell, B. 6988) Nature 311,525-527 23. Kitaguchi, N., Takahashi, Y., Tokushima, Y., Shiojiri, S., and Ito, H.(1988) Nature 3 1 1 , 530-532 24. Tanzi, R. E., McClatchey, A. I., Lamperti, E. D., Villa-Komaroff, L., Gusella, J., and Neve, R. L. (1988) Nature 31 1 , 528-530 25. Konig, G., Monning, U., Czech, C., Prior, R., Banati, R., Schreiter-Gasser, U., Bauer, J., Masters, C. L., and Beyreuther, K. (1992) J. Biol. Chem. 267,10804-10809 26. Innocenti, G. N., Koppel, H., and Clarke, S. (1983) Deu. Brain Res. 1 1 , 39-53 27. Perry, V. H., and Gordon, S. (1988) Trends Neurosci. 11,273-277 28. McGeer, P. L., Itagaki, S., and McGeer, E. G. (1988) Acta Neuropathol. 7 6 , " I
~~
55n-557 "".
29. Budka, H. (1989) Acta Neuropathol. 77,225-236 30. Weckerle, H., Linington, C., Lassmannn, H., and Meyermann, R. (1986) Trends Neurosci. J. 9 , 271-277 31. Popovic, M., Sarngadharan, M. G., Read, E., and Gallo, R. C. (1984) Science 224,497-500 32. Boyum, A. (1968) Scand. J. Clin. Lab. Inuest. 2 1 , Suppl. 97, 31-50 33. Banati, R. B., Hoppe, D., Gottmann, K., Kreutzberg, G. W., and Kettenmann, H. (1991) J. Neurosci. Res. 30,593-600 34. Laemmli, U. K. (1970) Nature 227,680-685 35. Wessel, D., and Flugge, U. I. (1984) Anal. Biochern. 138,141-143 36. Towbin, H., Strahlein, T., and Gordon, T. (1979) Proc. Natl. Acad. Sci. U. S. A. 76,4350-4354 37. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochern. 1 6 2 , 156-159 38. Messing, J. (1983) Methods Enzyrnol. 101,ZO-78 39. Golde, T. E., Estus, S., Younkin, L. H., Selkoe, D. J., and Younkin, S. G. (1992) Science 2 6 5 , 728-730 40. Haass, C., Hung, A.Y., and Selkoe, D. J. (1991) J. Neurosci. 11, 37833793 41. Siman, R., Card, J. P., Nelson, R. B., and Davis, L. G. (1989) Neuron 3 , 275-285 42. Banati, R. B., Rothe, G., Kreutzberg, G. W., and Valet, G. (1991) J. Neuropathol. Ap 1 NeurobioL 17,223-230 43. Estus, S., Golde $'E., Kunishita, T., Blades, D., Lowery, D.,Eisen, M., Usiak, M., Q;, X, Tabira, T., Greenberg, B. D., and Younkin, S. G. (1992) Science 2 5 5 , 726-728 44. McGeer, P. L., McGeer, E. G., Kawamata, T., Yamada, T., and Akiyama, H. (1991) Can. J. Neurol. Sci. I S , 376-379 45. Wisniewski, H. M., Wegiel, J., Wang, K. C., Kujawa, M., and Lach, B. (1989) Can. J.Neurol. Sct. 16. 535-542 46. Frei, K.; and Fontana, A (1989)Neuroimmune Networks: Physiology and Diseases (Goetzel, E. J., and Spetor, N. H., eds) pp. 127-136, Alan R.
__ .-.,
"_..
1 . i Tnr ~ ~ N e w Vnrk l_"l,
47. Springer, T. (1990) Nature 346,425-434 48. Kametani, F., Haga, S., Tanaka, K., and Ishii, T. (1990) J. Neurol. Sci. 9 7 , 43-52 49. Chen, M., and Yankner, B. A. (1991) Neurosci. Lett. 125,223-226 51. Breen, K. C., Bruce, M., and Anderton, B. H. (1991) J. Neurosci. Res. 2 8 , 90-100 52. Klier, F. G., Cole, G., Stallcup, W., and Schubert, D. (1990) Brain Res. 615,336-342 53. Narindrasorasak, S., Lowery, D., Gonzalez-DeWhitt, P., Poorman, R. A., Greenberg, B., and Kisilevsky, R. (1991) J. Biol. Chern. 2 6 6 , 1287612883 54. Bauer, J., Konig, G., Strauss, S., Jonas, U., Ganter, U., Weidemann, A., Monning, U., Masters, C. L., Volk, B., Berger, M., and Beyreuther, K. (1991) FEBS Lett. 282,335-340 55. Gaudernack, G., Leivestad, T., Ugelstad, J., andThorsby,E. (1986) J. Immunol. Methods 90,179-187
