Amyloid Peptide-Induced Cytokine and Chemokine Expression in THP-1 Monocytes Is Blocked by Small Inhibitory RNA Duplexes for Early Growth Response-1 Messenger RNA

Alzheimer’s disease (AD)³ is a progressive neurodegenerative disease, the most frequent cause of dementia affecting >5% of the population over the age of 65 years. The neuropathology of AD is characterized by the presence of numerous intraneuronal deposits of neurofibrillary tangles, senile plaques around reactive microglia, and progressive loss of neurons in the brain (1). The major component of senile plaques and cerebrovascular deposits is amyloid β (Aβ), consisting of 39–43 aa residues, formed by the proteolytic processing of amyloid precursor protein (2). The accumulation of Aβ is thought to be an early and perhaps necessary feature of AD (3). The predominant forms of Aβ are the 1–40 and 1–42 fragments. Soluble Aβ_1–40 is the major form of circulating Aβ, while the amyloidogenic Aβ_1–42 is the major constituent of senile plaques and is present in minor amounts in the circulation (4, 5, 6, 7). Recent studies (8) have shown that mean values of Aβ_1–42 in the plasma of AD patients and control subjects were 236 ng/ml (52 nM) and 38 ng/ml (8.4 nM), respectively. However, some AD patients had Aβ_1–42 as high as 658 ng/ml (146 nM) in the plasma (8). It is thought that the Aβ peptides in the AD brain are produced locally, although some studies indicate that circulating Aβ may contribute to the amyloid deposits in the AD brain (7, 9).

In AD and Aβ-related cerebral vascular disorders (cerebral amyloid angiopathy and hereditary cerebral hemorrhage with amyloidosis of the Dutch type), one finds not only the increased deposition of Aβ in the brain, but also an increased presence of monocytes/macrophages in the blood vessel walls and activated microglial cells in the brain parenchyma (1, 10, 11, 12). Early studies by Hickey and co-workers (13) showed that peripheral hemopoietic cells (e.g., monocytes) could cross the blood-brain barrier (BBB), and then undergo differentiation in to microglial cells in the brain parenchyma. Recent studies by Eglitis and Mezey (14) give further credence that microglial cells in the brain are probably derived from hemopoietic cells. These studies thus provided compelling evidence that hemopoietic cells can cross the BBB and act as progenitor cells for the microglia. In vivo studies show that Aβ can induce the activation and migration of monocytes across a rat mesenteric vascular bed (15), indicating that a similar phenomenon can occur in the brain vasculature.

We and others have shown (16, 17) that Aβ can induce the migration of monocytes across a monolayer of normal human brain endothelial cells that serves as a model of the BBB. However, it is unclear whether Aβ interaction with monocytes caused an inflammatory response, which could allow activated monocytes to migrate across the cerebral vasculature and cause neuronal dysfunction. The idea that inflammatory molecules produced by activated monocytes, microglial cells, and astrocytes may play a role in neuronal dysfunction and death is supported by in vitro studies (18, 19, 20, 21, 22, 23). Moreover, there is evidence that amyloid deposition initiates a microglial-mediated inflammatory response that culminates in neuronal loss and cognitive decline in AD (24). Both clinical trial and epidemiological data also show that anti-inflammatory therapies, such as use of nonsteroidal anti-inflammatory drugs reduce the incidence and progression of AD (23, 25). These studies indicate the importance of inflammation in the pathogenesis of AD.

The results presented in this study show that Aβ peptides (Aβ_1–40 and Aβ_1–42) at physiological concentrations, as found in the plasma of AD individuals (8), initiate cellular signaling in THP-1 monocytes and peripheral blood monocytes (PBM) to increase the gene expression of specific cytokines (TNF-α and IL-1β) and chemokines (monocyte chemoattractant protein-1 ((MCP-1), IL-8, and macrophage inflammatory protein-1β (MIP-1β)). The cellular signaling involves activation of tyrosine kinase and extracellular signal-regulated kinase (ERK-1 and ERK-2), but not of p38 mitogen-activated protein kinase (MAPK). The amyloid-induced activation of ERK-1/ERK-2 leads to activation of transcription factors, early growth response-1 (Egr-1) and AP-1, but not of NF-κB or CREB. We show that transfection of THP-1 cells with small interfering RNA duplexes containing 22 nt, specific for Egr-1 mRNA (Egr-1 small inhibitory RNA (siRNA)), blocks Aβ-induced expression of TNF-α, IL-1β, MCP-1, IL-8, and MIP-1β. This is the first report, to our knowledge, showing that inhibition of activation of transcription factor, specifically Egr-1; by Egr-1 siRNA can block the Aβ-mediated inflammatory response.

The THP-1 promonocytic cell line was obtained from American Type Culture Collection (Manassas, VA). Cells were cultured in RPMI 1640 containing 10% heat-inactivated FCS, 10 mM HEPES, 2 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. PBM were isolated from blood collected in EDTA as the anticoagulant as previously described (26). PD 98059, U0126, GF 109203X, AG 490, and Genistein were purchased from BIOMOL (Plymouth Meeting, PA). GW 5074, SB 203580, and SP 600125 were obtained from Tocris (Ellisville, MO). Anti-phospho-p42/44 MAPK Abs (E10, monoclonal) was purchased from Cell Signaling (Beverly, MA). Rabbit anti-ERK-1 Ab (SC-93), anti-phospho-Elk-1 Ab (SC-8406, monoclonal), rabbit anti-Egr-1 Ab (SC-110X), anti-phosphotyrosine Ab conjugated with HRP (SC-7020, monoclonal), rabbit anti-c-Fos Ab (SC-7202X), goat anti-c-Jun/AP-1 Ab (SC-45X), rabbit anti-NF-κβ p65 Ab (SC-372X), anti-NF-κβ p50 Ab (SC8414X, monoclonal), goat anti-SP-1 (SC-59X), and secondary Abs conjugated to HRP were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other reagents, unless otherwise specified, were purchased from Sigma-Aldrich (St. Louis, MO).

Amyloid peptides, Aβ_1–40 and Aβ_1–42, were custom synthesized at W. M. Keck facility at Yale University, purified, and characterized by analytical reverse phase HPLC, amino acid analysis, and laser desorption spectrophotometry as described previously (27). Both peptides were dissolved in endotoxin-free water (Sigma-Aldrich) at a concentration of 2 mg/ml, while Aβ_1–42 was aged for 7-days at 37°C. Both peptides were aliquoted and stored at 4°C. These peptide solutions were negative for endotoxin (<10 pg/ml) as determined by Limulus lysate test (Sigma-Aldrich). We determined the physical state of the amyloid peptide used in these experiments using thioflavin T fluorescence assay (28) and Far-UV CD spectra (29). We found that Aβ_1–40, which was dissolved in water at a concentration of 2 mg/ml (∼460 μM) and kept at 4°C for 30 days, showed significant amount of fibrillar form. However, Aβ_1–40, which was freshly prepared in water or dissolved in DMSO at a concentration of 2 mg/ml showed an absence of fibrils. However, Aβ _1–42 (2 mg/ml) when dissolved in water and kept at 37°C for 7 days showed maximal amount of fibril formation. We used nonfibrillar Aβ_1–40 and fibrillar Aβ_1–42 in the studies described herein.

THP-1 monocytes were treated with Aβ_1–40 peptide for various time periods, and total RNA was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA). RNase protection assays (RPA) were performed on total RNA extracted from THP-1 cells using custom-made multiprobe templates for TNF-α, IL-1β, RANTES, MIP-1β, MCP-1, and IL-8, and the housekeeping genes L-32 and GAPDH (BD PharMingen, San Diego, CA). Briefly, templates were labeled with [α-³²P]UTP using T7 RNA polymerase according to manufacturer’s protocol. Ten micrograms of total RNA was hybridized with ³²P-labeled probe (8 × 10⁵ cpm) for 12–16 h at 56°C. The contents were treated with RNase mixture (BD PharMingen), followed by extraction with phenol-chloroform. Protected mRNA hybrids were resolved on a 6% denaturing polyacrylamide-sequencing gel and exposed to x-ray film for 24 h. The intensity of bands corresponding to TNF-α, MIP-1β, IL-1β, MCP-1, IL-8, L-32, and GAPDH were analyzed using an α Imager 2000 gel documentation system (α Innotech, San Leandro, CA). Values were expressed as relative expression of mRNA normalized to the mean of L32 and GAPDH mRNA.

THP-1 cells (5 × 10⁶ cells/ml) were incubated in serum-free RPMI 1640 in the presence or the absence of Aβ_1–40 peptide (125 nM). At indicated periods (1–4 h) cells were removed by centrifugation at 500 × g for 10 min. The medium was collected and stored at −80°C until further use. The amount of secreted chemokines (MIP-1β, MCP-1, and IL-8) was assayed using specific Duo-Set ELISA development systems (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions.

For Western blot analysis, THP-1 cells were cultured in RPMI 1640 medium containing 10% FBS for 3–4 days. On the day of experiment, cells were pelleted and resuspended at 1 × 10⁶cells/ml in serum-free RPMI 1640 medium and incubated for an additional 3 h before treatment with Aβ_1–40 peptide (125 nM). Where indicated, THP-1 monocytes were incubated with pharmacological inhibitors for 30 min before Aβ_1–40 treatment. The medium was aspirated, and cells were lysed in RIPA buffer (1× PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM sodium orthovanadate, 10 μg/ml PMSF, and 1 μl/ml protease inhibitor cocktail). The concentration of protein in each sample was quantified by Bradford method (Bio-Rad, Hercules, CA). Ten micrograms of proteins were size fractionated in 10% SDS-PAGE gel and transferred to nitrocellulose membrane (Bio-Rad). Blots were probed with anti-phosphotyrosine Ab (Santa Cruz Biotechnology, Santa Cruz, CA). Activation of ERK-1/ERK-2 was assessed using a 1/1000 dilution of anti-phospho-p42/44 Ab (Cell Signaling Technology, Beverly, MA). HRP-conjugated secondary Abs were used to develop the membrane and visualization of bands was performed using Supersignal chemiluminescent substrate (Pierce, Rockford, IL). Blots were stripped and reprobed using a 1/1000 dilution of Abs against the p42/44 to normalize the protein loading. The intensity of bands was quantified utilizing α Imager 2000 gel documentation system (α Innotech).

Nuclear extracts were prepared from THP-1 cells according to the modified procedure of Dignam et al. (30). Briefly, 5 × 10⁶ cells were washed with cold PBS. Cells were resuspended in 400 μl of cell lysis buffer (10 mM HEPES (pH 7.9), 100 mM KCl, 1.5 mM MgCl₂, 0.1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF, 0.5% Nonidet P-40, and 1 μl/ml protease inhibitor cocktail (Calbiochem, La Jolla, CA)) and allowed to swell on ice for 30 min, followed by vigorous vortex mixing for 5–10 s. The homogenate was centrifuged in a microfuge at 10,000 × g for 30 s. Supernatant was discarded, and the nuclear pellet was resuspended in 50 μl of nuclear extraction buffer (10 mM HEPES (pH 7.9), 1.5 mM MgCl₂, 420 mM NaCl, 0.1 mM EGTA, 0.5 mM DTT, 5% glycerol, 0.5 mM PMSF, and 1 μl/ml protease inhibitor cocktail). The tube was mixed intermittently for 60 min. The nuclear extract was obtained by centrifuging at 10,000 × g for 10 min at 4°C.

The oligonucleotide used as probes were as follows: Egr-1, 5′-GGA TCC AGC GGG GGC GAG CGG GGG CGA-3′ and 3′-CCT AGG TCG CCC CCG CTC GCC CCC GCT-5′; AP-1, 5′-CGC TTG ATG ACT CAG CCG GAA-3′ and 3′-GCG AAC TAC TGA GTC GGC CTT-5′; NF-κβ, 5′-AGT TGA GGG GAC TTT CCC AGG C-3′ and 3′-TCA ACT CCC CTG AAA GGG TCC G-5′; and CREB 5′-AGA GAT TGC CTG ACG TCA GAG AGC TAG-3′ and 3′-TCT CTA ACG GAC TGC AGT CTC TCG ATC-5′, and were synthesized at Norris Cancer Center Microchemical core facility at University of Southern Calfornia. Probes were 5′ end labeled with 100 μCi of [γ-³²P]ATP using T4-polynucleotide kinase. The labeled single-stranded sense oligonucleotide probe was mixed with labeled antisense probe and incubated at 65°C for 5 min, followed by annealing at room temperature for 15 min. The DNA binding reaction mixture contained nuclear proteins (2–4 μg), ³²P-labeled double-stranded oligonucleotide probe (∼50,000 cpm), and 2 μg of poly(dI-dC). To demonstrate the specificity of DNA-protein interaction, a 50-fold excess of unlabeled double-stranded oligonucleotide probe was added. In supershift assays, nuclear extracts were preincubated for 20 min at room temperature with 2 μg of Ab specific to each transcription factor before the addition of radiolabeled probe. The DNA-protein complex was then size fractionated from the free DNA probe by electrophoresis in a 4% nondenaturing polyacrylamide gel. The gel was vacuum dried and exposed to x-ray film.

The 22-nt sequence of Egr-1 siRNA was derived from human Egr-1 mRNA sequence (GenBank accession no. GI: 5420378), and was targeted to the coding region 1237–1258 relative to the start codon of Egr-1 gene. The Egr-1 siRNA oligonucleotides were synthesized with following sequences: antisense Egr-1, 5′-AGG CAT ACC AAG ATC CAC TTG CT ATA GTG AGT CGT ATT A-3′; and sense Egr-1, 5′-CCG CAA GTG GAT CTT GGT ATG CT ATA GTG AGT CGT ATT A-3′, and were synthesized containing T7 promoter sequence, which is underlined. The scrambled Egr-1 siRNA (sc Egr-1 siRNA) was synthesized using the sense oligonucleotides (5′-TCG ACA ACA TCA CTG AGA TCG CTA TAG TGA GTC GTA TTA-3′) and antisense (5′-GAG CGA TCT CAG TGA TGT TGT CTA TAG TGA GTC GTA TTA-3′). The small RNA was generated by T7-oligonucleotide-directed in vitro synthesis as previously described (31). Briefly, sense or antisense Egr-1-T7 oligonucleotides (25 ng) were incubated separately in an in vitro transcription buffer containing a mixture of 0.4 mM cocktail (ATP, CTP, GTP, and UTP), T7 RNA polymerase (50 U), 40 mM Tris-HCl, pH 8.0, 8 mM MgCl₂, 2 mM spermidine-HCl, 25 mM NaCl, and 5 mM DTT for 10 min at 37°C. The contents of sense and antisense transcripts produced were mixed and incubated at 65°C for 5 min, followed by annealing for 15 min at room temperature. To the annealed reaction mixture were added 0.1 vol of 3 M sodium acetate, pH 5.2, and 2.5 vol of cold ethanol. The contents were kept at −70°C for 60 min, and RNA (Egr-1siRNA duplex) was pelleted and dissolved in 50 μl of RNase-free water. The amount of RNA was quantified by absorption at 260 nm.

THP-1 cells were transfected with either Egr-1 siRNA or scEgr-1 siRNA duplex using Lipofectamine. Briefly, Egr-1 siRNA (0.25 or 0.5 μg/ml) was incubated in 100 μl of serum-free DMEM containing 10 μl of Lipofectamine (Invitrogen) for 15 min as previously described (32). The complexes were added to the THP-1 cell suspension (2 × 10⁶ cells/ml) and incubated at 37°C in a tissue culture incubator with 5% CO₂ for 24 h. Then, the medium was aspirated and replaced with RPMI 1640 containing 10% FBS. After 48–72 h of transfection, cells were harvested and used for additional experiments.

Statistical analysis of the responses obtained from control and Aβ-treated THP-1 monocytes was conducted by one-way ANOVA using the Instat 2 (GraphPad, San Diego, CA) software program. The effects of inhibitors on Aβ-induced responses were analyzed by comparing the responses of THP-1 monocytes in the presence and the absence of inhibitor. Student’s test was used for multiple comparisons. A value of p < 0.05 was considered significant.

Previous studies have shown that treatment of THP-1 monocytes or PBM with Aβ at a micromolar concentrations (10–25 μM) resulted in secretion of cytokines IL-6 and TNF-α, and chemokines (16, 33, 34). Since our studies (17) showed that nanomolar concentrations of Aβ_1–40 (125 nM) and Aβ_1–42 (125 nM) were effective in mediating transmigration of monocyte across a monolayer of cultured human brain endothelial cells (17, 26), and submicromolar concentrations of amyloid peptide are observed in plasma of AD subjects (8), we studied the effect of Aβ over this nanomolar range. As shown in Fig. 1,A, treatment of THP-1 monocytes with Aβ_1–40 for 2 h resulted in increased mRNA expression of TNF-α, MIP-1β, IL-1β, MCP-1, and IL-8 in a dose (62.5–500 nM)-dependent manner compared with basal level of expression, as determined by RPA. However, the basal mRNA levels of RANTES and CCR2b remained unchanged in response to Aβ_1–40. As shown in Fig. 1,B, the optimal mRNA expression of these cytokines and chemokines occurred at a concentration of 125 nM Aβ_1–40. The fold increase in the mRNA expression of TNF-α (1.6), MIP-1β (2.7), IL-1β(2.0), MCP-1 (2.3), and IL-8 (1.7) in response to Aβ_1–40 (125 nM) is indicated in parentheses. As a positive control, LPS (100 and 500 ng/ml) treatment of THP-1 cells caused increased mRNA expression of TNF-α, MIP-1β, IL-1β, MCP-1, and IL-8 and down-regulation of CCR-2a and CCR-2b (Fig. 1 A).

The increase in mRNA expression of cytokines and chemokines induced by Aβ_1–40 was not due to contamination of synthetic Aβ_1–40 preparation with endotoxin, as Aβ_1–40 was dissolved in endotoxin-free water. Additionally, polymyxin B (5 μg/ml), an inhibitor of endotoxin, did not significantly inhibit the Aβ-induced mRNA expression of these cytokines and chemokines (data not shown), whereas polymyxin B (5 μg/ml) completely abrogated the LPS (100 ng/ml)-induced increase in the mRNA expression of these cytokines and chemokines (data not shown).

Fig. 2 A shows the effect of Aβ_1–40 (125 nM) treatment of THP-1 monocytes, as a function of time, on the mRNA expression profile of cytokines and chemokines. As shown, mRNA expression of TNF-α, MIP-1β, and IL-1β was optimal at 2 h, while the expression levels of MCP-1 and IL-8 were optimal at 1 h. After this optimal time period, mRNA levels of these cytokines and chemokines returned to basal levels.

Since Aβ treatment of THP-1 monocytes caused increased gene expression of both cytokines (TNF-α and IL-1β) and chemokines (MCP-1, IL-8, and MIP-1β), we examined the effect of pharmacological inhibitors, which have been previously shown to be specific for various kinases in the signaling pathway (35, 36, 37, 38). Fig. 3 shows mRNA profile of cytokines (TNF-α and IL-1β) and chemokines (MCP-1, IL-8, and MIP-1β) in THP-1 monocytes, which were preincubated for 30 min with the indicated pharmacological inhibitors in the absence and the presence of Aβ_1–40 (125 nM). As shown in Fig. 3, pretreatment of THP-1 cells with genistein (25 μg/ml), a protein tyrosine kinase inhibitor (35), inhibited Aβ-induced mRNA expression of TNF-α, IL-1β, MCP-1, and IL-8 by >90%, while expression of MIP-1β was reduced by ∼50%. A specific inhibitor of MAPK kinase (MAPKK/MEK), PD 98059 (10 μM) (36), completely inhibited Aβ-induced mRNA expression of TNF-α, IL-1β, MCP-1, and IL-8, while MIP-1β expression was reduced by ∼75%. Similarly, U0126 (150 nM), a specific inhibitor of MEK1/2 (39), abrogated Aβ-induced mRNA expression of TNF-α, IL-1β, MCP-1, IL-8, and MIP-1β. However, SB 203580 (1–2 μM), a selective p38 MAPK inhibitor (37), did not significantly affect the Aβ-induced mRNA expression of these cytokines and chemokines. Next, we examined the effect of SP 600125 (100 nM), a potent inhibitor of c-Jun N-terminal kinase (JNK) (38). As shown in Fig. 3 A, lane 8, SP 600125 reduced the expression of TNF-α, IL-1β, and IL-8 by >80%. However, expression of MIP-1β and MCP-1 was reduced by ∼60 and 50%, respectively. These results indicate that Aβ-induced mRNA expression of these cytokines and chemokines involves activation of ERK-1/ERK-2, but not p38 MAPK. Moreover, Aβ-mediated expression of these cytokines and chemokines also involves activation of JNK.

To delineate the cellular component, upstream of MAPK, involved in regulating the expression of these aforementioned cytokines and chemokines, we used wortmannin, a specific PI3-kinase inhibitor (37). Wortmannin (100–200 nM) did not affect the expression of these cytokines and chemokines (data not shown). Additionally, GF 109203X (20–40 nM), a protein kinase C inhibitor, was ineffective in inhibiting Aβ-induced mRNA expression of these cytokines and chemokines in THP-1 monocytes (data not shown). However, GW 5074 (20 nM), a potent and specific inhibitor of c-Raf (40), reduced by >95% the expression of TNF-α, IL-1β, and MCP-1, while expression of MIP-1β was inhibited by ∼60% (Fig. 3, lane 7). It is pertinent to note that GW 5074 further enhanced (150%) Aβ-induced expression of IL-8 (Fig. 3).

We determined whether interaction of amyloid peptide with PBM exhibited similar changes in cytokine and chemokine mRNA expression as observed in THP-1 monocytic cells. As shown in Fig. 4, treatment of PBM with Aβ_1–40 (125 nM) resulted in an increase in TNF-α, MIP-1β, IL-1β, MCP-1, and IL-8 mRNA expression as was observed in THP-1 monocytes. Moreover, pharmacological inhibitors, PD 98059 (MEK-1/2 kinase inhibitor) and GW 5074 (c-Raf kinase inhibitor), which reduced Aβ_1–40-induced expression of TNF-α, MIP-1β, IL-1β, and MCP-1 in THP-1 cells also reduced the expression in PBM. Furthermore, GW 5074, which does not inhibit IL-8 mRNA expression in THP-1 cells was also not effective in PBM (Fig. 4). It is pertinent to note that basal levels of IL-1β and IL-8 mRNA in PBM were much higher than those in THP-1 cells. Since amyloid peptide showed a similar profile of cytochemokine gene expression in both THP-1 cells and PBM, we used THP-1 monocytic cells as a model system for subsequent studies.

Since chemokines play a role in chemotaxis and Aβ was effective in increasing the gene expression of chemokines (MIP-1β, MCP-1, and IL-8), we measured the secretion of these chemokines in the Aβ-treated THP-1 monocytes. THP-1 monocytes were kept in serum-free medium for 2 h, followed by treatment with Aβ_1–40 (125 nM) for 1, 2, and 4 h. At the end of incubation, medium was collected for measurement of secreted chemokines by employing ELISA assay as described by the manufacturer (R&D Systems). Fig. 5,A shows the effect of Aβ_1–40 on the release of MIP-1β as a function of time. The data show an ∼3-fold increase in the secretion of MIP-1β at 2 h (1275 ± 83.47 pg/ml in Aβ-treated vs 435.22 ± 23.36 pg/ml in untreated THP-1 monocytes). Similarly, there was an ∼3-fold increase in the secretion of MCP-1 (837.36 ± 23.72 pg/ml in Aβ-treated vs 273.74 ± 27.8 pg/ml in untreated THP-1 monocytes) at 2 h (Fig. 5,B). IL-8 shows optimal secretion at 2 h in Aβ_1–40-treated THP-1 monocytes (1506 ± 9102.8 pg/ml) over its corresponding control (761.4 ± 62.6 pg/ml) as shown in Fig. 5 C.

Since Aβ_1–40-induced chemokine (MCP-1, IL-8, and MIP-1β) mRNA expression was attenuated by MAPK inhibitor (PD 98059, U0126), c-Raf kinase inhibitor (GW 5074), and JNK inhibitor (SP 600125), we examined the effects of these inhibitors on Aβ-induced secretion of these chemokines. As shown, these inhibitors reduced >80% of MIP-1β (Fig. 5,D) and MCP-1 (Fig. 5,E) secretion. A similar effect was observed with genistein. However, these inhibitors approximately inhibited >50% of Aβ-induced secretion of IL-8 (Fig. 5 F). The extent of inhibition of release of these chemokines by MAPK inhibitor (PD 98059 and U0126), c-Raf kinase inhibitor (GW 5074), and JNK inhibitor (SP 600125) was similar to the reduction of Aβ-induced transcription of these genes by these inhibitors.

Since Aβ-induced mRNA expression of cytokines and chemokines was inhibited by tyrosine kinase inhibitor, we determined the tyrosine phosphorylation pattern of proteins in THP-1 cells stimulated by nanomolar concentrations of Aβ_1–40 (125 nM). Previous studies (41) showed that micromolar (50 μM) concentrations of Aβ_25–35 caused an increase in protein tyrosine phosphorylation in THP-1 monocytes. As shown in Fig. 6,A, Aβ_1–40 (125 nM) caused increases in overall tyrosine phosphorylation of several proteins (∼42–44, 60, 80, and 120 kDa), which were inhibited by genistein, PD 98059, and SP 600125. However, SB 203580 (a p38 kinase inhibitor), wortmannin (PI3 kinase inhibitor), GF 1092903X (protein kinase C inhibitor), and AG 490 (JAK2 inhibitor) did not significantly affect tyrosine phosphorylation of proteins (Fig. 6 A).

We determined whether Aβ_1–40 stimulated protein tyrosine kinase in THP-1 monocytes caused downstream activation of members of the MAPK superfamily. As shown in Fig. 6,B, treatment of THP-1 monocytes with Aβ_1–40 (125 nM) for time periods ranging from 1–60 min, followed by Western blot analysis with an Ab to phosphorylated ERK-1/ERK-2 revealed increased phosphorylation of both ERK-1 and ERK-2 up to 30 min. The maximal increase in phosphorylation (∼3-fold) of ERK-1/ERK-2 occurred at 30 min in response to treatment of THP-1 monocytes with Aβ_1–40 (125 nM). The peak phosphorylation of ERKs (ERK-1/ERK-2) observed at 30 min declined at 60 min. To delineate which pathways are critical for Aβ_1–40-mediated activation of ERKs, we treated THP-1 monocytes with specific inhibitors of different signaling pathways before the addition of Aβ_1–40. In the presence of genistein, ERK phosphorylation in Aβ-stimulated THP-1 monocytes was completely inhibited, indicating that activation of tyrosine kinase was required for Aβ-induced ERK activation (Fig. 6,C). As expected on the basis of the capacity of Aβ to activate MEK, PD 98059, an inhibitor of MEK, completely abolished ERK phosphorylation in THP-1 monocytes. Moreover, SP 600125, a JNK inhibitor, reduced Aβ-induced ERKs phosphorylation by ∼60%. However, SB 203580, a p38 MAPK inhibitor, did not inhibit ERK phosphorylation, indicating that p38 MAPK is not involved in Aβ-induced activation of ERKs. Wortmannin, AG 490 (inhibitor of JAK 2), and GF 109203X did not inhibit Aβ-induced ERKs phosphorylation (Fig. 6 C).

Activation of ERK leads to its translocation into the nucleus, where it activates Elk-1 (42). Elk-1 is activated by phosphorylation at several different serine residues, the most critical of which appears to be serine 383. To examine whether Aβ-induced ERK activation was able to phosphorylate Elk-1 at serine 383, we used an Ab that specifically recognizes Elk-1 phosphorylated at serine 383. Aβ induced a time-dependent (30–240 min) increase in the phosphorylation of Elk-1 (Fig. 7,A). The same blot was stripped and reprobed with SP-1 Ab to normalize protein loading. Fig. 7 B shows densitometric analysis of the Western blot of Elk-1 phosphorylation. At 4 h, Elk-1 phosphorylation was ∼8-fold higher in Aβ_1–40-treated THP-1 monocytes (Aβ-treated, 3.67 ± 0.77 arbitrary units; control, 0.44 ± 0.05 arbitrary units) than the untreated THP-1 monocytes. There was no statistical difference (p < 0.05) in Elk-1 phosphorylation between 2 and 4 h.

Next, we determined whether Aβ-induced activation of ERKs and Elk-1 resulted in activation of transcription factor Egr-1, because Egr-1 is a target transcription factor of activated Elk-1 (43, 44). The involvement of Egr-1 in Aβ-mediated induction of cytokines and chemokines is likely, since some of these cytokines and chemokine genes have Egr-1 consensus sites in their promoters (45, 46). For this reason, we examined whether Aβ-induced Egr-1 DNA binding activity. As shown in Fig. 8,A, a time-dependent (0.5–4 h) increase in Egr-1 DNA binding activity was observed in nuclear extracts prepared from Aβ-treated THP-1 monocytes (Fig. 8,A, lanes 3–6). At 4 h there was a substantial decrease in Egr-1 DNA binding activity (Fig. 8,A, lane 6). As a positive control, LPS activated Egr-1 DNA binding activity in THP-1 monocytes (Fig. 8,A, lane 8) as previously observed (43). The data in Fig. 8,A show 8-fold increase in Egr-1 DNA binding activity in Aβ-treated THP-1 monocytes. The specificity of this interaction was determined by adding excess unlabeled Egr-1 probe to the binding reaction. The Egr-1 signal was completely reduced (Fig. 8,B, lane 4). Furthermore, Ab to Egr-1 supershifted the band corresponding to Egr-1 (Fig. 8,B, lane 5). As a negative control, Ab to SP-1 and NF-κβ (p65) failed to supershift the Egr-1 band (Fig. 8,B, lanes 6 and 7, respectively). Additionally, Genistein, PD 98059, and SP600125, which all inhibited Aβ_1–40-mediated cytokine and chemokine gene expression, also reduced Egr-1 DNA binding activity by ∼90% (Fig. 8 C).

As shown in Fig. 8,A, a time-dependent (0.5–4 h) increase in AP-1 DNA binding activity was observed in nuclear extracts prepared from Aβ-treated THP-1 monocytes (Fig. 9,A, lanes 3–6). At 4 h there was a substantial decrease in AP-1 DNA binding activity (Fig. 9,A, lane 6). As a positive control, LPS activated AP-1 DNA binding activity in THP-1 monocytes (Fig. 9,A, lane 8). The data in Fig. 9A show an ∼5-fold increase in AP-1 DNA binding activity in Aβ-treated THP-1 monocytes. The specificity of this interaction was determined by adding excess unlabeled AP-1 probes to the binding reaction. The AP-1 signal was completely reduced (Fig. 9,B, lane 4). Furthermore, Abs to c-Jun and c-Fos supershifted the band corresponding to AP-1 (Fig. 9,B, lanes 5 and 6). Fig. 9,C shows the effects of pharmacological inhibitors on AP-1 DNA binding activity. Both genistein and PD 98059 attenuated AP-1 DNA binding activity by 80 and 100%, respectively, (Fig. 9,C, lanes 4 and 5). Furthermore, SP 600125, a JNK/stress-activated protein kinase, completely inhibited AP-1 DNA binding activity (Fig. 9 C, lane 6).

Because Aβ_1–40 caused activation of MEK in THP-1 monocytes, which is also known to activate inactive NF-κB/Iκβ complex by causing phosphorylation of IκB, followed by dissociation of active NF-κB complex, we determined whether Aβ_1–40 resulted in the activation of NF-κB. Our data show that Aβ_1–40 does not cause a time-dependent (30–240 min) increase in NF-κB DNA binding activity (Fig. 10,A, lanes 3–6) above the basal level (Fig. 10,A, lane 2). However, LPS (100 ng/ml), as a positive control, increased by 5-fold the NF-κB DNA binding activity at 30 min (Fig. 10,A, lane 7). The specificity of NF-κB DNA binding complex was shown by supershift gel analysis in the presence of specific Ab to either p50 or p65 protein (Fig. 10,A, lanes 8 and 9). Similar data were obtained for CREB DNA binding activity (Fig. 10 B). These results show that Aβ_1–40 (125 nM) does not cause an increase in either NF-κB or CREB DNA binding activity in THP-1 monocytes.

Studies have shown that cytokine (TNF-α) and chemokine (MCP-1) genes have a consensus binding sites for Egr-1 within their promoter regions (46). We also show that Aβ increases Egr-1 DNA binding activity in THP-1 monocytes. Thus, we hypothesized that interference of Egr-1 protein production by small interfering RNA duplexes may interfere with Aβ-mediated Egr-1 DNA binding activity and concomitant inflammatory gene expression. The results presented in Fig. 11,A show that THP-1 cells transfected with increasing doses of Egr-1 siRNA show a dose-dependent reduction in Egr-1 protein levels in nuclear extracts, compared with nontransfected cells, in response to treatment with Aβ_1–40. As shown in Fig. 11,B, lane 2, and Fig. 11,C, lane 2, Aβ_1–40 increased Egr-1 DNA binding activity by 5-fold in THP-1 cells. The band corresponding to Egr-1 DNA complex was supershifted in the presence of Ab to Egr-1 (Fig. 11,C, lane 6). Furthermore, the Aβ-induced Egr-1 DNA binding activity was reduced in THP-1 cells transfected with increasing doses of Egr-1 siRNA. The maximal reduction (∼65%) of Egr-1 DNA binding activity was observed in THP-1 cells transfected with 500 ng/ml Egr-1 siRNA (Fig. 11,B, lane 4, and Fig. 11,C, lane 3). The specificity of Egr-1 siRNA in abrogating Aβ-mediated Egr-1 DNA binding activity was validated by using scrambled Egr-1 siRNA (sc Egr-1 siRNA). The data in Fig. 11 C, lane 4, shows that transfection of THP-1 cells with sc Egr-1 siRNA does not have any inhibitory effect on Aβ-induced Egr-1 DNA binding activity.

As shown in Fig. 11,D, treatment of Egr-1 siRNA-transfected THP-1 cells with Aβ_1–40 resulted in complete inhibition of IL-1β and IL-8, while the expression of MCP-1, TNF-α, and MIP-1β was reduced by ∼75%. However, the effect of Aβ_1–40 in inducing TNF-α, IL-1β, MCP-1, and IL-8 gene expression was not reduced in THP-1 cells transfected with scrambled Egr-1 siRNA (Fig. 11,D, lane 4). There was a modest (∼25%) reduction in MIP-1β mRNA expression (Fig. 11 D, lane 4). These results indicate that Aβ_1–40-mediated activation of TNF-α, MIP-1β, IL-1β, MCP-1, and IL-8 may be regulated either fully or partially by Egr-1.

Since soluble Aβ_1–40 increased the gene expression of cytokines and chemokines, we determined the effect of Aβ_1–42, a fibrillar form of amyloid peptide on THP-1 cells. As shown in Fig. 12,A, Aβ_1–42 treatment of THP-1 cells resulted in time-dependent (1–4 h) increase in the expression of TNF-α, MIP-1β, IL-1β, MCP-1, and IL-8, which were maximal at 1 h. Aβ_1–42 caused a time-dependent increase in phosphorylation of ERK-1 and ERK-2, which was maximal at 30 min (Fig. 12,B). Moreover, Aβ_1–42 showed strong increases in both Egr-1 and AP-1 DNA binding activity (Fig. 12,C). However, there was no significant increase in either NF-κB DNA binding activity or CREB DNA binding activity over the basal level (Fig. 12 C). These results indicate that both soluble Aβ_1–40 and the fibrillar form of Aβ_1–42 increase the gene expression of the same sets of cytokine and chemokine in THP-1 monocytes. Moreover, the extent of increase in the expression of these inflammatory genes occurs at physiological concentrations (125 nM) of both soluble and fibrillar forms of Aβ. As expected, treatment of Egr-1 siRNA-transfected THP-1 cells with Aβ_1–42 resulted in complete inhibition of IL-1β and IL-8 mRNA expression, while the expression of MCP-1, TNF-α, and MIP-1β was reduced by ∼75%. These data are similar to those observed with Aβ_1–40.

Previously (17) we have reported that Aβ_1–40 mediates the migration of monocytes across a monolayer of normal human brain endothelial cells that serves as a model of the BBB. Furthermore, we showed that interaction of Aβ_1–40 with the basolateral surface of brain endothelial cells derived from AD, but not from normal individual, resulted in an increased transmigration of monocytes from the apical to the basolateral direction (26), implying that the increased presence of amyloid peptide in the brain can cause signaling to favor migration of monocytes from the blood into the brain (26). Studies (13, 14) have shown that peripheral hemopoietic cells (e.g., monocytes) cross the BBB and differentiate into microglial cells in the brain parenchyma, suggesting that microglial cells arise from peripheral hemopoietic cells. Histological studies (11) have shown colocalization of Aβ peptide with activated microglial cells in AD brain. Numerous studies (16, 20, 33, 47, 48) have shown that amyloid peptide deposition in the brain provokes the microglial-mediated inflammatory response that contributes to neuronal apoptosis (18, 49) and the loss of memory that is characteristic of AD (24). Studies showed that interaction of fibrillar Aβ_1–42 at micromolar concentrations (10–40 μM) with either microglial cells or THP-1 monocytes, pretreated with LPS, elicited similar responses and resulted in increased secretion of TNF-α, IL-6, IL-8, and MIP-1β (28). However, the amounts of circulating amyloid peptides in plasma of AD subjects have been observed to be in the submicromolar range (8), thus suggesting that studies should be undertaken at these physiological concentrations.

In the present study we report that both soluble and fibrillar amyloid peptides, Aβ_1–40 and Aβ_1–42, respectively, at nanomolar concentrations (60–125 nM), increase the gene expression of cytokines (TNF-α and IL-1 β) and chemokines (MIP-1β, MCP-1, and IL-8) in THP-1 monocytes. Similar results were observed in PBM when treated with Aβ_1–40. The increase in expression of these inflammatory genes occurred at early time points, i.e., at 1–2 h, diminished by 4 h, and was completely abolished by 24 h. We next established Aβ_1–40-induced cellular signaling cascade pathways that lead to the increased expression of these cytokines and chemokines in THP-1 monocytic cells.

We show that Aβ_1–40-induced expression of cytokines (TNF-α and IL-1β) and chemokines (MCP-1 and IL-8) is inhibited >90% by genistein (a protein tyrosine kinase inhibitor), PD 98059 (a MEK inhibitor), and U0126 (an inhibitor of MEK1/2). However, SB 203580 (a p38 MAPK inhibitor), wortmannin (a PI3 kinase inhibitor), and GF.109203X (a protein kinase C inhibitor) had no significant effect on the expression of these cytokines and chemokines. These results indicate that Aβ_1–40-induced signaling leading to the expression of these cytokines and chemokines presumably involves downstream activation of protein tyrosine kinase and MAPK, but not p38 MAPK and PI3 kinase/Akt. This is supported by data showing that Aβ_1–40 causes phosphorylation of tyrosine residues in a subset of proteins, ranging in M_r from 40–180 kDa. Previous studies (18) showed that Aβ_1–40 and Aβ_25–35 at micromolar concentrations (50–60 μM) caused tyrosine phosphorylation of a subset of proteins in this range. Moreover, our studies show that Aβ_1–40 causes time-dependent increases in the phosphorylation of ERK-1 and ERK-2, which were completely inhibited by the pharmacological inhibitors, genistein and PD 98059, but not by wortmannin, SB203580, and GF 109203X. Furthermore, we observed that SP 600125, a specific inhibitor of JNK, reduced >60% the Aβ_1–40-induced phosphorylation of ERK-1/ERK-2. These results indicate that the Aβ-induced signaling cascade involves activation of protein tyrosine kinases, ERKs, and stress-activated protein kinases, also termed JNK, but not p38 MAPK, as illustrated in Fig. 13. Activation of p38 MAPK has been shown to occur in response to LPS (50) in THP-1 monocytes. However, the putative receptor (e.g., receptor for advance glycation end product, scavenger receptor) involved in the nonfibrillar and fibrillar forms of Aβ-mediated signaling in monocytes and microglia is controversial (22, 28, 41, 51).

Since activation of ERKs lead to its translocation into the nucleus where it activates Elk-1 (42), we show that Aβ_1–40 also causes a time-dependent (0.5–4 h) sustained increase in Elk-1 phosphorylation in THP-1 monocytes. Recent studies by Mackman and co-workers (43) show that LPS also induces phosphorylation of Elk-1 in monocytes.

Egr-1 is one of several target transcription factors of activated Elk-1 (43, 44). The promoter regions of TNF-α have putative binding sites for transcription factors Egr-1, AP-1, NF-κβ, and CREB (52, 53). The promoter regions of MCP-1 have binding sites for transcription factors AP-1 and NF-κβ (54). At present, to our knowledge, the promoter region of human MIP-1β has not been characterized. We examined whether nonfibrillar Aβ_1–40 interaction at nanomolar concentrations (125 nM) with THP-1 monocytic cells activated these transcription factors. We showed that Aβ_1–40 caused a time-dependent (30–240 min) increase in Egr-1 DNA binding activity, which was maximal at 60 min. Furthermore, we showed that Aβ_1–40 activates transcription factor AP-1, but had relatively no significant effect on the activation of NF-κB and CREB. We also observed that fibrillar Aβ_1–42 activated AP-1 and Egr-1, but not NF-κB and CREB. However, under the same conditions, LPS caused strong activation of NF-κβ in THP-1 monocytes. It is pertinent to note that both Aβ_1–40 and Aβ_25–35 at micromolar concentrations (50–60 μM) have been shown to cause activation of NF-κB in THP-1 monocytes (18). Under these conditions, there may have been increased production of reactive oxygen species, resulting in activation of NF-κB (55, 56).

Our results indicate that Aβ_1–40-induced gene expression of TNF-α, IL-1β, IL-8, and MCP-1 in THP-1 monocytes is probably regulated by activation of transcription factors Egr-1 and AP-1, but not NF-κB and CREB. This is further supported by data showing that inhibitors of upstream kinases, such as protein tyrosine kinases, ERK-1/2 kinases, and JNKs completely (∼90%) reduced Aβ_1–40-induced Egr-1 and AP-1 DNA binding activities, but not basal activity. These data correlate with the effect of these inhibitors in attenuating the Aβ_1–40-mediated gene expression of TNF-α, IL-1β, IL-8, and MCP-1, indicating either a direct or a causal effect of Egr-1 and AP-1 in transcriptional regulation of these cytochemokine genes. Although the pharmacological inhibitors used in our and others studies have been shown to be specific for specific kinases, their effect may not be specific as purported; hence, the use of dominant negative expression plasmids for signaling cascade kinases are warranted.

To address the role of Egr-1 in the regulation of transcription of TNF-α, IL-1β, IL-8, and MCP-1, we used siRNA duplexes for Egr-1 mRNA to reduce the corresponding protein and/or protein-dependent activity in THP-1 cells. The siRNA duplexes (21–23 nt) targeted to specific genes have been successfully introduced into mammalian cells in culture to reduce mRNA and protein expression (31, 32, 57). We show that transfection of THP-1 cells with Egr-1 siRNA reduced Egr-1 protein content by ∼60%. However, transfection of THP-1 cells with scrambled Egr-1 siRNA did not reduce Egr-1 protein content (data not shown), indicating the specificity of Egr-1 siRNA in these studies. The Egr-1 siRNA, but not scrambled Egr-1 siRNA, transfected THP-1 cells also showed reduced (∼60%) Egr-1 DNA binding activity (by EMSA) in response to Aβ_1–40. Furthermore, we showed that Aβ_1–40-induced mRNA expression of IL-1β and IL-8 was completely abrogated in Egr-1 siRNA-transfected THP-1 monocytic cells. To our knowledge, the promoter region of IL-1β does not contain any Egr-1 putative binding sites, yet studies by Pinsky and co-workers (58) showed that administration of antisense Egr-1 oligodeoxyribonucleotide to rats after lung transplantation reduced the expression of IL-1 β. Moreover, it has been shown that Egr-1 knockout mice fail to express IL-1β in response to ischemia/reperfusion compared with wild-type mice (45). Thus, we suggest that IL-1β expression may be regulated by Egr-1-dependent pathways (59). It should be pointed out that IL-8 promoter has AP-1, NF-κB, and CREB (60) binding sites. Although the IL-8 promoter gene does not have an Egr-1 binding site, it has been shown that activation of AP-1 is involved in the regulation of IL-8 expression (61). We hypothesize that Egr-1 may be regulating the activation of AP-1. It has been shown that activation of c-Jun is inhibited by a dominant negative Egr-1, indicating that AP-1 is downstream of Egr-1 (62, 63). Our data indicate that Aβ_1–40-mediated increased expression of IL-1β and IL-8 results from activation of Egr-1-dependent processes.

Moreover, Aβ_1–40-induced mRNA expression of TNF-α, MCP-1, and MIP-1β was reduced to ∼75% in Egr-1 siRNA-transfected THP-1 monocytic cells, indicating that these genes also use Egr-1 transcription factor, although a contribution of the transcription factor AP-1 in the activation of these genes cannot be ruled out (52, 54). Studies have shown that TNF-α promoter contains the binding sites for Egr-1, AP-1, NF-κB, and CREB (53, 64). At present, relatively little is known about the promoter of MIP-1β. The promoter for MCP-1 contains AP-1 and NF-κB DNA binding sites (54). The search of the database for MCP-1 gene and enhancer region (GI: 10933860) revealed a putative binding site, GCGGGGGAGG (nt 4278–4287), which corresponds to the Egr-1 binding site found in the M-CSF gene (59). Further studies of the promoter analysis of these genes will increase our understanding of how amyloid peptide increases their transcription.

In conclusion, our study shows that both soluble (nonfibrillar) and fibrillar amyloid peptides, which are present in AD brain at submicromolar concentrations, cause increased expression of cytokines (TNF-α and IL-1β) and chemokines (MCP-1, IL-8, and MIP-1β) in monocytes. We show that both Aβ_1–40- and Aβ_1–42-mediated cellular signaling cascades involve downstream activation of tyrosine kinases and ERKs and activation of transcription factors Elk-1 and Egr-1 (Fig. 13). The activation of JNKs and that of the transcription factor AP-1 are thought to occur in parallel (Fig. 13). We have provided evidence that the siRNAs approach, such as Egr-1 siRNA, functions effectively to silence the expression of Aβ_1–40-induced inflammatory genes, which have been implicated in the progression of AD. These studies provide a novel rational therapeutic approach to ameliorate the inflammation-induced progression of AD.

We thank Dr. Stanley Tahara for critically reading the manuscript. We appreciate the help of Drs. Ralf Langen and Sajith Jayasinghe in analyzing the nonfibrillar and fibrillar forms of amyloid peptides using Far-UV CD spectra and fluorescence spectroscopy.