Amyloid Precursor Protein Mediates Proinflammatory Activation of Monocytic Lineage Cells*

Alzheimer’s disease is a progressive neurodegenera- tive disorder characterized by extracellular deposition of (cid:1) -amyloid (A (cid:1) ) peptide containing neuritic plaques. A (cid:1) peptides are proteolytically derived from the mem-brane-bound amyloid precursor protein (APP). Al- though the function of APP is not entirely clear, previous studies demonstrate that neuronal APP colocalizes with (cid:1) 1 integrin receptors at sites of focal adhesion, suggesting that APP is involved in mediating neuronal process adhesion. Integrin-dependent adhesion is also a well-characterized component of immune cell proinflammatory activation. Using primary mouse microglia and the human monocytic cell line, THP-1, we have be-gun investigating the role of APP in integrin-dependent activation. Co-immunoprecipitation studies demonstrate that APP is recruited into a multi-receptor signal- ing complex during (cid:1) 1 integrin-mediated adhesion of monocytes. Stimulation induces a subsequent, specific recruitment of tyrosine phosphorylated proteins to APP, including Lyn and Syk. Antibody cross-linking of cell surface APP leads to a similar response characterized by activation and recruitment of tyrosine kinases to APP as well as subsequent activation of mitogen-acti-vated protein kinases and increased proinflammatory protein levels. These data demonstrate that APP can act as a proinflammatory receptor in monocytic lineage cells and provide insight into the contribution of this protein to the inflammatory conditions described in Alz- heimer’s disease.

Amyloid precursor protein (APP) 1 is a ubiquitously expressed integral membrane protein from which the 1-40 and 1-42 residue ␤-amyloid (A␤) peptides are proteolytically cleaved (1). The longer, more insoluble peptide, A␤1-42, is a marked component of the extracellular neuritic plaques characteristic of Alzheimer's disease (2). The physiologic role for APP independent of the production of the A␤ peptides remains to be elucidated.
Although only a fraction of the pool of cellular APP localizes to the plasma membrane, the overall structure of the protein suggests that APP may function as a receptor or growth factor (3). APP is an integral transmembrane glycoprotein that exists as one of three major splice variants consisting of either 695, 751, or 770 amino acid residues (4). These isoforms are composed of a large glycosylated extracellular region, a single membrane-spanning domain, and a short, highly conserved cytoplasmic tail (1). Interestingly, the cytoplasmic domain of APP contains a well-defined consensus motif found in tyrosine kinase receptors, NPXY, implicating a role in signal transduction that is reinforced by the ability of multiple adaptor proteins including Fe65 and X11 to interact with this specific motif on neuronal APP (5)(6)(7). The phosphotyrosine-binding domains of Fe65 and X11 bind the phosphotyrosine binding motif (Y 682 ENPTY 687 ) on the intracellular domain of APP. However, there is no evidence that phosphorylation of this domain is required for the interaction to occur (8,9). The adaptor protein Shc has also been identified as an APP-interacting protein; however, this interaction is promoted by tyrosine phosphorylation of the APP phosphotyrosine-binding motif (10,11). These data suggest that APP is capable of serving as a docking molecule in membrane proximal signaling events. It has also been demonstrated that neuronal APP colocalizes with ␤ 1 integrins at point contacts, suggesting a possible role in adhesion (12).
Furthermore, it has been demonstrated that APP binds directly to extracellular matrix molecules, particularly collagen type I (13).
Integrins are a family of heterodimeric cell surface receptors composed of an ␣ and ␤ subunit that are expressed in a diverse group of cell types. There are 16 identified ␣ and 8 ␤ subunits that combine to form at least 22 different ␣␤ heterodimeric integrin receptors. The specific combination of subunits dictates the binding specificity as well as overall function of the integrin receptor (14,15). Integrins function to modulate cellular adhesion and use a characteristic tyrosine kinase-mediated activation pathway to alter cellular response (16 -18). A common feature of integrin behavior is their ability to form macromolecular signaling complexes with other cell surface receptors (19 -21). ␤ 1 integrin-dependent adhesion stimulates a well-characterized tyrosine kinase-based activation response in monocytic lineage cells, inducing increased expression of several inflammatory mediator genes (22)(23)(24). Interestingly, inflammatory activation of these cells increases localization of APP to the plasma membrane, suggesting APP participates in adhesion-mediated activation of this cell type (25,26). To answer this question, we investigated the role of APP in ␤ 1 integrindependent adhesion and activation of the human monocytic cell line, THP-1, and primary mouse microglia.
Tissue Culture-THP-1 cells are a monocytic cell line derived from peripheral blood of a human with acute monocytic leukemia commercially available from the American Type Culture Collection (Manassas, VA). THP-1 cells were grown in RPMI 1640 (Life Technologies, Inc., Rockville, MD) containing 10% heat-inactivated fetal bovine serum (U.S. Biotechnologies, Inc., Parkerford, PA), 5 mM Hepes, and 1.5 g/ml penicillin/streptomycin/neomycin. Microglia were derived from the brains of postnatal day 2 C57Bl/6J mice. Briefly, cerebral cortices were isolated and trypsinized for 15 min. Digestion was terminated by adding tissue to DMEM/F12 media containing 10% fetal bovine serum and 5% horse serum (U.S. Biotechnologies, Inc.). Cells were triturated to obtain a single-cell suspension and plated into 75-cm 2 flasks. Media were replaced the next day, and cells were fed every 5 days. At ϳ14 days in vitro, microglia were harvested by rapid shaking for 30 min on a reciprocal shaker.
Cell Stimulation-THP-1 monocytes were removed from normal growth medium into serum-free RPMI 1640 and incubated at 37°C for 15 min before stimulation. Tissue culture wells were coated with type I rat tail collagen and allowed to dry in a laminar flow hood. THP-1 cells were added to either tissue culture plastic alone or to wells coated with collagen at a density of 3 ϫ 10 6 cells/ml. Cells were stimulated for either 15 min, 30 min, 60 min, or 24 h. In inhibition studies, cells were incubated at 37°C for 30 min with tyrosine kinase inhibitor, PP1 (5 M), before stimulation.
Generation of Fab Antibody Fragments-The monoclonal 22C11 antibody was digested at 37°C for 24 h in the presence of 0.02 M cysteine and 0.1 M papain. Digestion was terminated with the addition of 0.03 M iodoacetamide. Fab fragments were separated on a 10% nonreducing polyacrylamide gel and identified by colloidal Coomassie staining. The band was removed from the gel, and electro-elution was performed to recover the Fab fragments. The identity of these fragments was confirmed by Western blotting for APP.
APP Cross-linking-To cross-link cell surface APP, THP-1 cells or microglia were removed from normal growth medium into serum-free RPMI 1640 and incubated at 37°C for 15 min. Cells were either unstimulated (control) or stimulated with anti-APP antibody, 22C11 (1 g/ml), monovalent Fab APP antibody fragment (1 g/ml), or mouse IgG 1 , isotype control (1 g/ml). Cells were stimulated for 5 min or 24 h at 37°C. Total cell lysates and immunoprecipitates were prepared as described below.
FIG. 2. APP is incorporated in a multi-receptor signaling complex with ␤ 1 integrin. THP-1 cells were plated for 30 min on tissue culture plastic (c), type I collagen (col), 0.1 mg/ml fibronectin (fib), 0.1 mg/ml laminin (lam), or 0.05 mg/ml poly-L-lysine (lys). Cells were lysed in 1% Triton X-100 buffer, and APP was immunoprecipitated. A, cell lysates were resolved by 7% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, and anti-ERK2 antibody (loading control). B, immunoprecipitates were resolved by 7% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, and anti-APP antibody. C, THP-1 cells were plated for 30 min on tissue culture plastic alone (c) or on type I collagen (col). Cells were lysed in 1% Triton X-100 buffer, and either APP or ␤ 1 integrin was immunoprecipitated. Immunoprecipitates were resolved by 7% SDS-PAGE and Western blotted using anti-␤ 1 integrin antibody or anti-APP antibody. Antibody binding was visualized by chemiluminescence.
concentrations were quantitated by the method of Bradford (27). Proteins were resolved by 7 or 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes for Western blotting. Western blots were blocked in TBS-T (10 mM Tris, pH 7.4, 100 mM NaCl, and 0.1% Tween 20) containing 3% BSA for 15 min and then incubated overnight at 4°C in primary antibodies. Blots were washed three times in TBS-T, followed by incubation for 1 h with horseradish peroxidase-conjugated secondary antibodies in TBS-T containing 5% nonfat dried milk. The blots were washed three times in TBS-T, followed by detection with enhanced chemiluminescence (Pierce). In some instances, blots were stripped in 0.2 N NaOH, 5 min, 25°C.
Immunoprecipitation-For co-immunoprecipitation, cells were lysed in ice-cold Triton lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 1 mM Na 3 VO 4 , 10 mM NaF, 1 mM EDTA, 1 mM EGTA, 0.2 mM phenylmethylsulfonyl fluoride, and 1% Triton). Lysates were vortexed and then incubated on ice for 15 min, followed by pulse sonication. Cells were centrifuged 10 min at 4°C to remove insoluble material. Primary antibody (1 g/mg protein) was added and incubated 4 h at 4°C. Protein A/G beads (35 l) were added and incubated overnight at 4°C. Beads were washed three times with lysis buffer, and immunoprecipitates were resolved and Western blotted as described above.
Small Interference RNA (siRNA) Transfection-THP-1 cells were transfected with APP siRNA (2 ϫ 10 6 cells; 2 g of siRNA) using the appropriate Nucleofector program as described by the manufacturer (Amaxa Inc., Gaithersburg, MD). Cells were transfected with either individual siRNA duplexes or the combined pool of duplexes. Cells were stimulated by adhesion to collagen (as described above) 24 h after transfection or lysed to determine APP expression.

RESULTS
Collagen type I is the preferred substrate for a subset of ␤ 1 subunit containing integrin heterodimers (28 -30). ␤ 1 integrin ligation reportedly stimulates activation of cytoplasmic tyrosine kinases and regulates gene expression in monocytic lineage cells (22,24,31). To characterize a tyrosine kinase activation response in our system, we determined that adhesion of THP-1 monocytes to type I collagen stimulated a time-dependent increase in protein phosphotyrosine levels with a maximal stimulation at 30 min (Fig. 1A). Collagen adhesion stimulated increased phosphorylation, indicative of activation, of the p38 mitogen-activated protein kinase (MAPK) (Fig. 1B). However, adhesion-mediated activation did not stimulate activation of the ERKs nor JNKs (Fig. 1, C and D). It is important to point out that control cells display a slight increase in tyrosine phosphorylated protein levels and increased phosphorylation of p38, which is likely attributable to the progressive settling of the cells on the tissue culture plastic (Fig. 1, A and B). The profile of increased protein phosphotyrosine levels stimulated by collagen adhesion was different from that stimulated by other ␤ 1 ligands, fibronectin, and laminin, as well as the nonspecific cationic ligand poly-L-lysine ( Fig. 2A) (32). Additionally, only collagen-dependent adhesion stimulated recruitment of tyrosine phosphorylated proteins to APP (Fig. 2B). Moreover, collagen adhesion stimulated formation of a multi-receptor complex including cell surface APP and ␤ 1 integrin subunits (Fig.  2C). These data suggest that APP participates in the tyrosine kinase-dependent activation response of ␤ 1 integrin receptors in monocytes, particularly those using type I collagen as a ligand. Reportedly, the potential ␣␤ combinations that can function as collagen receptors include the receptors composed of the ␤ 1 subunit noncovalently linked to either ␣ 1 , ␣ 2 , ␣ 3 , ␣ 10 , or ␣ 11 subunit (29).
To further characterize the contribution of APP to ␤ 1 integrindependent activation, we identified tyrosine phosphorylated proteins recruited to APP during stimulation (Fig. 3A). Previous studies demonstrated that the non-receptor tyrosine kinases Lyn and Syk are active components of integrin-mediated signaling responses in monocytic cells (24,33). Co-immunoprecipitation experiments verified that Lyn and Syk were recruited to APP upon collagen binding and displayed increased levels of tyrosine phosphorylation, indicative of enzyme activa-tion (Fig. 3, A-C). Protein tyrosine kinases are often receptor proximal factors involved in the initial stage of a series of events leading to activation of cytoplasmic serine/threonine kinases and subsequent transcriptional regulation (34). The tyrosine kinase-initiated activation response mediated by ␤ 1 integrin-dependent adhesion of monocytic cells is a well-described event in the induction of immediate-early genes characteristic of monocytic differentiation including interleukin-1␤ (IL-1␤), interleukin-8 (IL-8), and tumor necrosis factor-␣ (35). As expected, persistent exposure of monocytes to collagen resulted in the acquisition of a reactive phenotype including an increase in cyclooxygenase-2 (COX-2), inducible nitric oxide synthase, IL-1␤, and CD36 protein levels (Fig. 4). To determine whether tyrosine kinase activation was required for subsequent activation of the p38 MAPK pathway and increased proinflammatory protein levels, we treated THP-1 cells with a FIG. 4. Adhesion-mediated activation stimulates acquisition of a reactive phenotype. THP-1 monocytes were plated on tissue culture plastic alone or on type I collagen for the indicated times (in hours). Cells were lysed in RIPA buffer, separated by 7% SDS-PAGE, and Western blotted using anti-CD36 antibody, anti-inducible nitric oxide synthase (␣-iNOS) antibody, anti-COX-2 antibody, anti-IL-1␤ antibody, and anti-ERK2 antibody (loading control). Antibody binding was visualized by chemiluminescence. specific inhibitor of Src family kinases, PP1, and evaluated the downstream effects (36). Pretreatment of THP-1 cells with PP1 before plating on collagen prevented the increased activation and recruitment of the tyrosine kinases to APP and subsequent activation of p38 MAPK (Fig. 5A). Importantly, kinase inhibition also prevented the stimulated increase in COX-2 protein levels (Fig. 5B), which served as a representative indicator of reactivity, demonstrating that Src family kinases such as Lyn are critical initiating factors in acquisition of a reactive phenotype. To further define a role for APP in integrin-mediated proinflammatory activation of monocytic lineage cells, we used siRNA to significantly reduce APP expression (Fig. 5C). We determined that APP is a critical component of adhesion-mediated increased p38 MAPK activity and increased COX-2 protein levels, indicating that APP is involved in acquisition of a reactive phenotype (Fig. 5, D and E).
Our findings demonstrate that APP has receptor-like properties upon ␤ 1 integrin engagement; however, we predicted that APP may function as a receptor independent of ␤ 1 integrin ligation. We tested this hypothesis by stimulating THP-1 cells with an antibody directed against the extracellular domain of APP, clone 22C11, to simulate ligand binding (37). Direct crosslinking of cell surface APP on THP-1 cells stimulated a rapid increase in tyrosine phosphorylated proteins qualitatively similar to that induced by collagen adhesion (Fig. 6A). In addition, co-immunoprecipitations verified that Lyn and Syk were activated and recruited to APP upon antibody cross-linking similar to adhesion-dependent activation (Fig. 6B). However unlike collagen-dependent activation, antibody-dependent APP dimerization stimulated increased activation of not only p38 MAPK but also ERKs and JNKs (Fig. 6A). These data demonstrate fundamental differences in APP contribution to monocytic activation, depending upon its recruitment to ␤ 1 integrin macromolecular complexes or its ability to act as an independent receptor. Importantly, we demonstrated that proinflammatory activation was dependent on APP dimerization because the monovalent Fab antibody fragment was not sufficient to stimulate increased tyrosine phosphorylated proteins or increased MAPK activity (Fig. 6C).
We next determined whether APP dimerization stimulated FIG. 5. Kinase recruitment to APP is required for adhesion-stimulated acquisition of a reactive phenotype. THP-1 cells were either untreated or treated for 30 min with PP1 (5 M) or dimethyl sulfoxide vehicle. Cells were plated for 30 min or 24 h on tissue culture plastic alone or type I collagen. A, cells were lysed at 30 min with 1% Triton X-100 buffer, and APP was immunoprecipitated. Cell lysates were resolved by 7 or 10% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, anti-phospho-p38 antibody, and anti-p38 antibody. Immunoprecipitates were resolved by 7% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, anti-Lyn antibody, and anti-APP antibody. B, cells were lysed at 24 h with RIPA buffer. Cell lysates were resolved by 7% SDS-PAGE and Western blotted with anti-COX-2 antibody anti-ERK2 antibody (loading control). C, THP-1 cells were either not transfected (c) or transfected with individual APP siRNA duplexes or pooled duplexes using the Amaxa Nucleofector transfection system. Cells were lysed at 24 h with RIPA buffer. Cell lysates were resolved by 7% SDS-PAGE and Western blotted with anti-APP antibody. D, non-transfected cells were unstimulated (control) or stimulated by adhesion to collagen (col), or cells were stimulated 24 h after transfection with siRNA by adhesion to collagen (col APP Ϫ ). Cells were lysed at 15 or 30 min in RIPA buffer, separated by 7% SDS-PAGE, and Western blotted using anti-phosphotyrosine antibody, 4G10, anti-phospho-p38 antibody, and anti-p38 antibody. E, cells were stimulated 24 h after transfection with siRNA by adhesion to collagen. Cells were lysed at 24 h in RIPA buffer, separated by 7% SDS-PAGE, and Western blotted with anti-COX-2 antibody and anti-ERK2 antibody (loading control). Antibody binding was visualized by chemiluminescence. acquisition of a reactive phenotype similar to that induced by ␤ 1 integrin engagement. Dimerization stimulated an increase in COX-2 and IL-1␤ protein levels in the monocytes after 24-h treatment (Fig. 7). However, in contrast to collagen-stimulated activation, direct APP cross-linking did not stimulate increased inducible nitric oxide synthase or CD36 protein levels. These data again illustrate a difference between adhesion-dependent and APP dimerization-dependent activation of monocytic cells and define the specificity of APP independent-mediated changes in a cellular proinflammatory phenotype.
To confirm that initial tyrosine kinase activity was critical for subsequent activation of the MAPK pathways and the phenotypic changes, THP-1 cells were pretreated with PP1 to inhibit Lyn activation before cross-linking. Kinase inhibition prevented the stimulated increase in tyrosine phosphorylated proteins and recruitment of Lyn and Syk to APP as well as the increase in p38 MAP kinase activity (Fig. 8A). Surprisingly, ERK and JNK activation were unaffected by kinase inhibition, demonstrating a divergence in the signaling response downstream of APP (Fig. 8A). Similar to adhesion-mediated activation, tyrosine kinase inhibition also prevented the acquisition of a reactive phenotype represented by the lack of a stimulated increase in COX-2 protein levels (Fig. 8B). FIG. 7. APP cross-linking stimulates increased COX-2 and IL-1␤ protein levels. THP-1 cells were unstimulated (control) or stimulated for indicated times (in hours) with anti-APP, 22Cll clone (1 g/ml), or mouse IgG 1 (negative control, 1 g/ml). Cells were lysed with RIPA buffer, separated by 7% SDS-PAGE, and Western blotted with anti-COX-2 antibody, anti-IL-1␤ antibody, anti-inducible nitric oxide synthesis (␣-iNOS) antibody, anti-CD36 antibody, and anti-ERK2 antibody (loading control). Antibody binding was visualized by chemiluminescence. FIG. 6. APP dimerization stimulates a tyrosine kinase-dependent signaling response in THP-1 cells. THP-1 cells were unstimulated (c) or stimulated for 5 min with anti-APP, 22C11 clone (1 g/ml), monovalent Fab APP antibody fragment (1 g/ml), or mouse IgG 1 (negative control, 1 g/ml). Cells were lysed in 1% Triton X-100 buffer, and APP was immunoprecipitated. A, cell lysates were resolved by 7 or 10% SDS-PAGE and Western blotted using anti-phosphotyrosine antibody, 4G10, anti-phospho-p38 antibody, anti-p38 antibody, anti-phospho-ERK antibody, anti-ERK2 antibody, antiphospho-JNK antibody, and anti-JNK antibody. B, immunoprecipitates were resolved by 7% SDS-PAGE and Western blotted using anti-phosphotyrosine antibody, 4G10, anti-Lyn antibody, anti-Syk antibody, and anti-APP antibody. C, cell lysates were resolved by 7 or 10% SDS-PAGE and Western blotted using antiphosphotyrosine antibody, 4G10, antiphospho-p38 antibody, anti-p38 antibody, anti-phospho-ERK antibody, anti-ERK2 antibody, anti-phospho-JNK antibody, and anti-JNK antibody. Antibody binding was visualized by chemiluminescence.
Although the primary focus of this report is characterizing the contribution of APP to monocyte activation, we presumed that APP participates in specific microglial activation paradigms as well. For comparison, we demonstrate that crosslinking cell surface APP on microglial cells also stimulated a rapid increase in protein phosphotyrosine levels, activation of Lyn, activation of ERKs (Fig. 9, A and B), and increased COX-2 protein levels (Fig. 9C). Importantly, APP dimerization did not induce activation of JNKs or p38 MAPK as observed in monocytes (Fig. 9A). These data demonstrate that APP is capable of acting as an independent receptor responsible for tyrosine kinase-mediated proinflammatory activation of microglia. However, the specific activation response diverges from that observed in THP-1 monocytes, suggesting differences in the resultant phenotype. DISCUSSION These data define a novel role for APP in the proinflammatory activation of monocytic lineage cells. We demonstrate that FIG. 8. Tyrosine kinase inhibition prevents cross-link-stimulated kinase recruitment to APP and subsequent increase in COX-2 protein levels. THP-1 cells were either untreated or treated for 30 min with PP1 (5 M) or dimethyl sulfoxide (DMSO) vehicle. Cells were then either unstimulated or stimulated for 5 min or 24 h with anti-APP, 22C11 clone (1 g/ml), or mouse IgG 1 (negative control, 1 g/ml). A, cells were lysed at 5 min with 1% Triton X-100 buffer, and APP was immunoprecipitated. Cell lysates were resolved by 7 or 10% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, anti-phospho-JNK antibody, anti-JNK antibody, anti-phospho-ERK antibody, anti-ERK2 antibody, anti-phospho-p38 antibody, and anti-p38 antibody. Immunoprecipitates were resolved by 7% SDS-PAGE and Western blotted with anti-phosphotyrosine antibody, 4G10, anti-Lyn antibody, and anti-APP antibody. B, cells were lysed at 24 h with RIPA buffer, separated by 7% SDS-PAGE and Western blotted with anti-COX-2 antibody and anti-ERK2 antibody (loading control). Antibody binding was visualized by chemiluminescence. FIG. 9. APP cross-linking stimulates a tyrosine kinase-dependent signaling response and stimulates increased COX-2 protein levels in primary mouse microglia. A, primary microglia were unstimulated (control) or stimulated for the indicated times with anti-APP, 22C11 clone (1 g/ml), or mouse IgG 1 (negative control, 1 g/ml). Cells were lysed with RIPA buffer, separated by 7 or 10% SDS-PAGE, and Western blotted with anti-phosphotyrosine antibody, 4G10, antiphospho-ERK antibody, anti-ERK2 antibody, anti-phospho-p38 antibody, anti-p38 antibody, anti-phospho-JNK antibody, and anti-JNK antibody. B, cells were unstimulated (c) or stimulated for 5 min with anti-APP, 22C11 clone (1 g/ml), or mouse IgG 1 (negative control, 1 g/ml). Cells were lysed with RIPA buffer, separated by 7% SDS-PAGE, and Western blotted with anti-phosphotyrosine antibody, 4G10, antiphospho-Lyn antibody, and anti-Lyn antibody. C, cells were lysed at 24 h with RIPA buffer, separated by 7% SDS-PAGE, and Western blotted with anti-COX-2 antibody and anti-ERK2 antibody (loading control). Antibody binding was visualized by chemiluminescence. ␤ 1 integrin ligation stimulates increased tyrosine kinase activity, recruitment of tyrosine kinases to APP, and a resultant reactive phenotype, indicating that APP is an important signaling molecule in adhesion-mediated activation of THP-1 monocytes. We confirmed the role for APP in adhesion-mediated proinflammatory activation through the use of APP siRNA. Decreasing APP expression does not affect initial tyrosine phosphorylation of cytoplasmic proteins, consistent with previously demonstrated data in which integrin-mediated activation is sufficient to stimulate tyrosine kinase activity (24,33,38). However, downstream activation of p38 MAPK and subsequent increased COX-2 protein levels are dependent on APP expression. This suggests that APP does not have a role in regulating adhesion-mediated membrane proximal tyrosine kinase activity, but it does have a role in the subsequent signaling and proinflammatory phenotype. Presumably, recruitment of these initial activated tyrosine kinases to APP is critical for the propagation of the proinflammatory signaling response, therefore decreasing APP expression is sufficient to attenuate the proinflammatory signaling cascade.
Importantly, we also demonstrate that cell surface APP dimerization stimulates a collectively similar activation of monocytic lineage cells. Previous work has identified a hydrophobic region on the ectodomain of cell surface APP that is proposed to play an important functional role, such as dimerization or ligand binding (3). Our data support the idea that there is a biological ligand for APP and indicate that receptor ligation stimulates a series of signaling events, ultimately leading to transcriptional changes and an inflammatory state in monocytic lineage cells. Interestingly, the resultant phenotype differs slightly between APP dimerization and adhesion-mediated activation, which further supports the proposal that APP has a specific ligand and consequently an independent function as a cell surface receptor.
We report the non-receptor tyrosine kinases Lyn and Syk as novel proteins that interact with APP during activation of this putative receptor. It remains to be determined if the interaction between APP and these tyrosine kinases is a phosphorylation-independent event as described by the interaction between neuronal APP and the cytoplasmic proteins Fe65, X11, and murine DAB-1 (6,8,39). It may be more probable that the interaction emulates that demonstrated between C-terminal fragments of APP and the adaptor proteins Shc and Grb2 in which phosphorylation of Y 682 within the Y 682 ENPTY 687 motif promotes the interaction of these proteins (10,11). Additionally, this same tyrosine residue is phosphorylated by a constitutively active form of the non-receptor tyrosine kinase Abl and serves as a docking site for the Src homology 2 (SH2) domain of Abl itself (40). The nature of protein-protein interactions involving non-receptor tyrosine kinases of the Src family is welldescribed and involves the Src homology 2 domain of these proteins interacting with a phosphorylated tyrosine residue on the coupled protein. The exact nature of the interaction between APP, Lyn, and Syk will be addressed to a greater extent in future work.
The involvement of APP in proinflammatory activation of peripheral blood cells reveals that APP has an important role in modulating immune function. This is a previously undescribed function for APP, and it may provide important insight into potential immune system dysfunction in patients with autosomal dominant forms of Alzheimer's disease. APP dimerization in primary murine microglia also stimulates an inflammatory state, although the signaling pathway diverges slightly from that demonstrated in peripheral monocytes. This difference may be a reflection of heterogeneity between monocytes and microglia, between species, or between primary cells and cell lines.
In conclusion, these data demonstrate a novel function for APP as a cell surface receptor that mediates proinflammatory activation of monocytic lineage cells. Moreover, these data support the idea that there is a yet unidentified APP ligand and demonstrate the significance of this ligand in stimulating a tyrosine kinase-dependent signaling cascade in monocytes and microglia that incites the production of inflammatory mediators such as COX-2 and IL-1␤. The physiological consequences of this inflammatory state in vitro have yet to be studied, although it is probable that in vivo the release of these inflammatory mediators may lead to activation of cells in the immediate environment, such as endothelial cells or astrocytes. Importantly, tyrosine kinase inhibition prevents the increased COX-2 protein levels and could possibly serve as a critical regulatory point in this inflammatory pathway.