Monocyte p110 phosphatidylinositol 3-kinase regulates ... - CiteSeerX

Monocyte p110␣ phosphatidylinositol 3-kinase regulates phagocytosis, the phagocyte oxidase, and cytokine production Jimmy S. Lee,*,† William M. Nauseef,‡ Alireza Moeenrezakhanlou,*,† Laura M. Sly,§,ⱍⱍ Sanaa Noubir,*,† Kevin G. Leidal,‡ Jamie M. Schlomann,‡ Gerald Krystal,§,ⱍⱍ and Neil E. Reiner*,†,¶,1 † Vancouver Coastal Health Research Institute (VCHRI) and Departments of *Medicine (Division of Infectious Diseases), ¶Microbiology and Immunology, Faculties of Medicine and Science, and §Pathology, University of British Columbia, and ⱍⱍTerry Fox Laboratory, British Columbia Cancer Agency, Vancouver, British Columbia, Canada; and ‡ Inflammation Program and Department of Medicine, University of Iowa, Veterans Administration Medical Center, Iowa City, Iowa, USA

Abstract: Mononuclear phagocytes are critical modulators and effectors of innate and adaptive immune responses, and PI-3Ks have been shown to be multifunctional monocyte regulators. The PI-3K family includes eight catalytic isoforms, and only limited information is available about how these contribute to fine specificity in monocyte cell regulation. We examined the regulation of phagocytosis, the phagocyte oxidative burst, and LPSinduced cytokine production by human monocytic cells deficient in p110␣ PI-3K. We observed that p110␣ PI-3K was required for phagocytosis of IgG-opsonized and nonopsonized zymosan in differentiated THP-1 cells, and the latter was inhibitable by mannose. In contrast, p110␣ PI-3K was not required for ingestion serum-opsonized zymosan. Taken together, these results suggest that Fc␥R- and mannose receptor-mediated phagocytosis are p110␣-dependent, whereas CR3-mediated phagocytosis involves a distinct isoform. It is notable that the phagocyte oxidative burst induced in response to PMA or opsonized zymosan was also found to be dependent on p110␣ in THP-1 cells. Furthermore, p110␣ was observed to exert selective and bidirectional effects on the secretion of pivotal cytokines. Incubation of p110␣-deficient THP-1 cells with LPS showed that p110␣ was required for IL-12p40 and IL-6 production, whereas it negatively regulated the production of TNF-␣ and IL-10. Cells deficient in p110␣ also exhibited enhanced p38 MAPK, JNK, and NF-␬B phosphorylation. Thus, p110␣ PI-3K appears to uniquely regulate important monocyte functions, where other PI-3K isoforms are uninvolved or unable to fully compensate. J. Leukoc. Biol. 81: 1548 –1561; 2007. Key Words: PI-3K 䡠 mannose 䡠 Fc␥ receptor 䡠 complement receptor 3 䡠 lentivirus 䡠 siRNA 1548

Journal of Leukocyte Biology Volume 81, June 2007

INTRODUCTION Mononuclear phagocytes are important regulators and effectors of the innate and acquired immune responses [1], and hence, there is significant interest in understanding how they are regulated. Extensive research has highlighted an important role for phosphoinositides (PIs) in monocyte cell regulation. The 3⬘-PI metabolites produced by the PI-3K family of lipid kinases are known to be involved in regulating numerous monocyte activities including phagocytosis [2], oxidative burst [3], and TLR-mediated cytokine secretion [4 – 6]. An important research objective is to develop an understanding of how specificity is achieved in PI-3K signaling for these diverse biological functions. The PI-3Ks constitute a family of lipid kinases, which phosphorylate the hydroxyl group of the inositol ring of PIs at the 3⬘ position. The 3⬘-PI metabolites, produced as a result, are known to be involved in regulating a multitude of cellular events, such as mitogenic responses, differentiation, apoptosis, cytoskeletal organization, membrane traffic along the exocytic and endocytic pathways (reviewed in ref. [7]), and various other aspects of monocyte function [8 –10]. Part of this diversity of cellular control is differentially regulated by distinct PI-3K isoforms [11–13]. In vivo, Class I PI-3K isoforms phosphorylate principally phosphatidylinositol (4,5)-bisphosphate to generate phosphatidylinositol (3,4,5)-trisphosphate (PIP3) [14, 15]. Class I PI-3Ks are divided further into two subclasses (IA and IB), both of which are known to be activated by cell surface receptors. Mammalian Class IA PI-3Ks are heterodimers consisting of a regulatory subunit (p85␣, p85␤, p55␥, or other splice variants) coupled with a p110 (␣, ␤, or ␦ isoforms) catalytic subunit [16 –18]. Through their Src homology 2 (SH2)

1 Correspondence: Division of Infectious Diseases, University of British Columbia, Rm. 452D, 2733 Heather St., Vancouver, BC, Canada, V5Z 3J5. E-mail: [email protected] Received September 12, 2006; revised February 20, 2007; accepted February 22, 2007. doi: 10.1189/jlb.0906564

0741-5400/07/0081-1548 © Society for Leukocyte Biology

domain-containing p85 subunits, Class IA PI-3Ks are recruited to and are activated by cell surface receptors with intrinsic protein tyrosine kinase activity or to receptors coupled to cytoplasmic tyrosine kinases. Several lines of evidence have implicated a role for PI-3K in phagocytosis. PI-3K inhibitors wortmannin and LY294002 inhibit the uptake of IgG-coated RBC or latex beads by macrophages and neutrophils [2, 19 –21]. Class IA PI-3K has also been implicated in regulating phagocytosis, based on its association with Syk in Fc␥R-mediated uptake [22]. Phagocytosis mediated by complement receptor 3 (CR3) in murine macrophages has also been shown to be impaired under conditions of PI-3K inhibition [23]. The mechanism by which PI-3K regulates CR3-mediated phagocytosis is not clear, and it is not known which isoform of PI-3K is involved. Professional phagocytes have critical roles in host defense, and the production of microbicidal oxidants is a key component of innate immunity. The generation of microbicidal oxidants by monocytes and neutrophils requires the activation of a multiprotein complex, NADPH oxidase. The importance of the phagocyte oxidative burst is illustrated by a genetic disease in humans, chronic granulomatous disease (CGD), characterized by severe, recurrent bacterial and fungal infections [24]. Genetic defects in CGD patients result in absent or defective oxidase subunits [24]. Cells from these patients are normal in terms of phagocytosis, chemotaxis, and degranulation but are unable to generate reactive oxygen intermediates [25]. Assembly of the NADPH-dependent oxidase requires PI-3K activity and 3⬘-PIs [25, 26]. PI-3K serves at least two important roles in oxidase activation, including recruitment and localization of p47phox and p40phox to the membrane and activation of p47phox [3]. The isoform of PI-3K, which regulates activation of the oxidase, is not known. Mononuclear phagocytes detect infection through pattern recognition receptors, most prominently the TLR family (reviewed in ref. [27]). Class IA PI-3K has been shown to be linked to the innate immune response through TLRs in macrophages, dendritic cells (DC), endothelial cells, and B cells [6, 28 –30]. Various studies have examined the role of PI-3K in the secretion of proinflammatory cytokines induced by bacterial LPS through TLR4 as well as other TLR and the related IL-1 receptor (IL-1R) [31, 32]. The results obtained are not entirely consistent, as positive and negative influences have been reported [29, 30, 33, 34]. A limitation common to most of the studies that investigated PI-3K and TLR4 signaling was that the roles of individual PI-3K isoforms were not examined directly. Given the evidence of nonredundancy in function amongst some p110 PI-3K catalytic subunits [e.g., p110␣ or p110␤ knockouts (KOs) were embryonically lethal], it may not be possible to assign a function to these PI-3K isoforms using conventional KO approaches. In the interest of overcoming some of these limitations and gaining insight to fine specificity of PI-3K cell regulation in monocytes, we report the results of studies which investigated cytokine secretion, phagocytosis, and oxidase activation in p110␣-deficient THP-1 cells.

MATERIALS AND METHODS Reagents RPMI 1640, DMEM, HBSS, penicillin/streptomycin, and 1 M HEPES solution were from Stem Cell Technologies (Vancouver, BC, Canada). PMSF anti-BSA antibody (B-7276), human IgM (I-8260), LPS from Escherichia coli O111:B4 (L-3012), cytochrome c from horse heart (C-7752), superoxide dismutase (SOD; S-8160), mannan (M-7504), PMA (P-1585), and wortmannin were obtained from Sigma-Aldrich (Oakville, ON, Canada). Antibodies to human p110␦ and LY294002 were from Calbiochem (San Diego, CA, USA). Antibody to human p110␣ (Clone 19) was from BD Biosciences PharMingen (Mississauga, ON, Canada). Antibodies to human p110␤ and actin were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Anti-p85-N-SH3 antibody was from Upstate Biotechnology (Lake Placid, NY, USA). Retinal pigment epithelial (RPE)-conjugated anti-CD11b antibody and RPE-conjugated isotype-matched control antibodies were from Caltag Laboratories (San Francisco, CA, USA). Antiphospho-p38 (Thr180/Tyr182), phospho-Erk (Thr202/Tyr204), phospho-JNK (Thr183/Tyr185), phospho-NF-␬B p65 (Ser536), and phosphoAkt (Thr308) antibodies were from Cell Signaling Technology (Beverly, MA, USA). Anti-GAPDH antibody was from Research Diagnostics (Concord, MA, USA). HRP-conjugated antirabbit, antimouse, and antigoat secondary antibodies were from Cedarlane Laboratories (Hornby, ON, Canada). Zymosan A and BSA were from MP Biomedicals, LLC (Irvine, CA, USA).

Cell lines The promonocytic cell lines THP-1 [American Type Culture Collection (ATCC), Manassas, VA, USA] were cultured in RPMI 1640 supplemented with 10% FBS (Life Technologies, Burlington, ON, Canada), 2 mM L-glutamine, penicillin (100 U/ml), and streptomycin (100 ␮g/ml). Cultures were maintained without exceeding 0.5 ⫻ 106 cells/ml. 293T human embryonic kidney (HEK) cells were also from ATCC and were cultured in DMEM, supplemented with 10% FBS, 2 mM L-glutamine, penicillin (100 U/ml), streptomycin (100 ␮g/ml), and 20 mM HEPES.

Lentivirus packaging, titration of lentiviral vectors, transduction of target cells PI-3K p110␣ isoform was silenced in THP-1 cells by lentiviral-delivered short hairpin (sh)RNA as described previously [35].

Western blot analysis Cells were washed once with PBS, lysed in boiling lysis buffer (1% SDS, 50 mM Tris, pH7.4, 0.15 M NaCl, 1 mM NaF, 10 mM PMSF, 1 mM sodium orthovanadate, 1 mM EDTA) for 5 min, and passed through a 27-gauge needle. Lysates were cleared by centrifugation at 12,000 g for 1 min, and protein concentration was determined using Bio-Rad DC protein assay. Equal amounts of protein were separated by 7.5% SDS-PAGE before transfer to nitrocellulose membranes, which were blocked with 5% skim milk or 3% BSA in TBST for 1 h at room temperature, depending on the antibody. Primary and secondary antibodies were used according to the manufacturer’s instructions, followed by detection with ECL technique (Amersham Biosciences, Piscataway, NJ, USA).

Cytokine measurement THP-1 cells at 1 ⫻ 106/ml were stimulated by serum-opsonized LPS at 100 ng/ml for the indicated times at 37°C, 5% CO2. Cultures were subsequently centrifuged at 400 g for 10 min at 4°C, and cell-free supernatants were collected and stored at –70°C until use. TNF-␣, IL-6, IL-12p40 (BD Biosciences PharMingen), and IL-10 (eBiosciences, San Diego, CA, USA) in culture supernatants were measured using sandwich ELISA, using paired cytokine-specific mAb, according to the manufacturer’s instructions.

Qualitative RT-PCR Total RNA was prepared using RNeasy kit (Qiagen, Valencia, CA, USA), and RT-PCR was performed using the Superscript First-Strand synthesis system for first-strand cDNA synthesis (Life Technologies, Grand Island, NY, USA). Taq DNA polymerase was from Fermentas (Burlington, ON, Canada). The primer sequences, number of cycles, and annealing temperature used for the various

Lee et al. PI-3K p110␣ gene silencing in monocytes

1549

cytokines are: TNF-␣, TCT CGA ACC CCG AGT GAC A (forward), GGC CCG GCG GTT CA (reverse), 35 cycles, 55°C; IL-6, ATG AAC TCC TTC TCC ACA AGC GC (forward), GAA GAA GCC CTC AGG CTG GAC TG (reverse), 35 cycles, 67°C; IL-10, ACG GCG CTG TCA TCG ATT (forward), TTG GAG CTT ATT AAA GGC ATT CTT C (reverse), 37 cycles, 56°C; IL-12p40, GGA TGC CGT TCA CAA G (forward), CCC ATT CGC TCC AAG A (reverse), 30 cycles, 60°C; actin, TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA (forward), CTA GAA GCA TTG CGG TGG ACG ATG GAG GG (reverse), 31 cycles, 55°C. An initial denaturation for 5 min at 94°C was followed by a cycle of 94°C for 1 min, the desired annealing temperature for 50 s, and 74°C for 1 min for extension, and the number of cycles depended on the primer used. The end of the cycle was followed by 10 min at 72°C. PCR products were run on 1% agarose gel in Tris-acetate-EDTA buffer, visualized with ethidium bromide.

Phagocytosis assay For each well, 0.5 ⫻ 106 THP-1 cells were differentiated with 10 ng/ml PMA for 48 h on 18 mm circular coverslips. Cells were then washed gently with HBSS twice and rested in fresh RPMI for 4 h at 37°C. Zymosan particles were unopsonized or opsonized with pooled human serum for 30 min at 37°C or coated with human IgG for 30 min at 37°C prior to use. Zymosan particles were added at a multiplicity of 5:1 or 10:1, and phagocytosis was allowed to proceed for 2 h at 37°C. In mannan-blocking experiments, mannan (at a final concentration of 0.1 mg/ml) was coincubated with cells during the final 15 min of the 4-h rest period. Fresh mannan (0.1 mg/ml) was also included with the zymosan mixture during the 2-h period of phagocytosis. All treatments were done on duplicate coverslips. Supernatants were discarded, and coverslips were fixed with formalin for 15 min, followed by mounting media. One hundred cells per coverslip were counted, and the average ingested particle:cell ratio and percentage of cells ingesting particles were calculated.

Superoxide assay Superoxide assays were performed as described previously by measuring SOD-inhibitable reduction of ferricytochrome c [36, 37] but with several modifications. Briefly, 0.5 ⫻ 106 THP-1 cells were differentiated overnight with 10 ng/ml (1.6 nM) PMA. Prior to stimulation, all cells were washed with HBSS twice and rested in RPMI for 4 h at 37°C. Zymosan particles were opsonized or not with pooled human serum for 30 min at 37°C prior to use. Cells were untreated or stimulated with 1 ␮M PMA or zymosan (20:1) in assay buffer (5.5 mM glucose, 79 ␮M cytochrome c) in the presence or absence of 30 ␮g/ml SOD. Each sample was mixed well and incubated at 37°C. Absorbance readings at 550 nm were done at 30 min poststimulation on a BioRad SmartSpec. PBS was used as blank control.

Statistical analysis For comparison of two treatment groups, a two-tailed t test was performed. A P value ⬍0.05 was considered significant. For three or more treatment groups, one-way ANOVA was performed on each group and followed by Tukey test for multiple comparisons. A P value ⬍0.05 was considered significant. All statistics and graphs were performed using GraphPad Prism software, Version 4.0 (GraphPad Software, San Diego, CA, USA).

RESULTS Effects of p110␣ silencing on phagocytosis of zymosan To investigate whether p110␣ regulates phagocytosis of zymosan particles, we used THP-1 cells deficient in p110␣ as a

Fig. 1. Phagocytosis of nonopsonized zymosan, IgG-opsonized zymosan, and serum-opsonized zymosan by PMA-differentiated THP-1 cells deficient in PI-3K p110␣ subunit. (A) Western blot analysis of Class IA PI-3K p110 catalytic subunit isoforms (␣, ␤, and ␦) and p85 regulatory subunit in THP-1 cells transduced with (HR-p110␣3, HR-p110␣1) or mock-transduced. Actin was used as protein-loading control. (B) THP-1 cells transduced with HR-p110␣1 (shRNA control) or HR-p110␣3 (p110␣ knockdown) were differentiated with PMA for 48 h and rested in RPMI for 4 h before exposure to nonopsonized zymosan (Zym), IgG-opsonized zymosan (IgG Zym), or serum-opsonized zymosan (Op Zym) for 2 h. In mannan-blocking experiments, mannan was added in the final 15 min of the rest period and during the 2-h phagocytosis period. Phagocytosis of zymosan was diminished significantly in p110␣ knockdown cells compared with control cells at a multiplicity of 5:1 and 10:1 (P⫽0.005 and P⫽0.0074, respectively, two-tailed paired t test, n⫽11). Phagocytosis of opsonized zymosan was not significantly (P⬎0.05) different in p110␣ knockdown cells compared with control cells (n⫽11). Phagocytosis of IgG-opsonized zymosan was significantly attenuated in p110␣-deficient cells (P⬍0.05, n⫽3). The presence of mannan significantly attenuated phagocytosis of zymosan in control cells (P⬍0.05, n⫽4) but not that of serum-opsonized zymosan. Error bars indicate SD.

1550


http://www.jleukbio.org

result of silencing by lentiviral-delivered small interfering (si)RNA (Fig. 1A) [35]. THP-1 cells were differentiated with PMA for 48 h before being exposed to zymosan opsonized with human serum, human IgG, or not opsonized. Cells deficient in p110␣ had reduced phagocytosis of unopsonized zymosan (P⬍0.05, n⫽11) but not serum-opsonized zymosan (n⫽11) compared with control cells at either particle:cell ratio (5:1 or 10:1; Fig. 1B). Like nonopsonized zymosan, phagocytosis of IgG-opsonized zymosan was significantly attenuated (P⬍0.05, n⫽3) in p110␣-deficient cells compared with controls. On average, the reduction in phagocytosis of zymosan and IgG zymosan in p110␣-deficient cells was 76.5% and 60%, respectively, compared with control cells. As the mannose receptor (MR) is another phagocytic receptor present on THP-1 cells [38], and the MR has been known to bind zymosan like CR3 [39], there was ambiguity as to whether the p110␣-dependent phagocytosis of zymosan observed in our system was mediated by CR3 or MR. It is interesting that preincubation with mannan reduced the phagocytosis of unopsonized zymosan to levels similar to p110␣-deficient cells, and it had no effect on serumopsonized zymosan (Fig. 1B). Binding experiments revealed that this decrease in phagocytosis in the presence of mannan correlated with a decrease in binding of unopsonized zymosan to the cell, and binding of serum-opsonized zymosan was not affected (data not shown).

Effect of p110␣ silencing on activation of the phagocyte oxidase by PMA and opsonized zymosan To examine the role of p110␣ in the oxidative burst, soluble and particulate agonists were used. THP-1 cells deficient in p110␣ or transduced with control shRNA were differentiated with low-dose PMA (1.6 nM) for 20 h and rested in fresh

medium for 4 h before incubation with high-dose PMA (1 ␮M), serum-opsonized zymosan, or nonopsonized zymosan (Fig. 2). As expected, PMA induced robust superoxide production by control cells, and cells deficient in p110␣ were impaired significantly (Fig. 2A) with a mean reduction of 76% (P⬍0.001). In response to opsonized zymosan, wild-type cells similarly produced significant superoxide compared with unstimulated cells (P⬍0.001). In contrast, opsonized zymosanstimulated THP-1 cells deficient in p110␣ had a significant reduction in superoxide production (Fig. 2B) with a mean decrease of 54% (P⬍0.01). In contrast to PMA or opsonized zymosan, no significant production of superoxide was detected in control cells when stimulated by nonopsonized zymosan (Fig. 2C). These results indicate that p110␣ was required for superoxide production in response to the soluble agonist PMA and the particulate agonist opsonized zymosan.

THP-1 cells deficient in PI-3K p110␣ showed dysregulation of LPS-induced cytokines To examine the role of p110␣ in regulating cytokine production, THP-1 cells deficient in p110␣ or control cells were incubated with LPS. Cytokine ELISAs were performed on supernatants collected at 5 h and 18 h postexposure to LPS. THP-1 p110␣-deficient cells showed enhanced production of TNF-␣ (Fig. 3A) and IL-10 (Fig. 3B) at both time-points compared with control cells. In contrast, LPS-induced levels of IL-12 and IL-6 were diminished significantly in p110␣-deficient cells (Fig. 3, C and D, respectively).

LPS-induced cytokine production in THP-1 cells is mediated through TLR4 To verify that LPS signaling in this model was through TLR4, THP-1 cells were preincubated for 1 h at 37°C with neutral-

Fig. 2. Class IA p85/p110␣ PI-3K is required for PMA- and opsonized zymosan-induced oxidase activation. THP-1 cells deficient in p110␣ or transduced with control shRNA were differentiated overnight with 10 ng/ml (1.6 nM) PMA and rested in RPMI for 4 h prior to stimulation. (A) PMA-induced superoxide production. THP-1 cells were stimulated or not with 1 ␮M PMA for 30 min. In response to PMA stimulation, cells deficient in p110␣ were reduced significantly in superoxide production compared with control cells (post-ANOVA Tukey test, P⬍0.001). Untreated control cells were not significantly different from stimulated, p110␣deficient THP-1 cells (post-ANOVA Tukey test, P⬎0.05, n⫽11). (B) Opsonized zymosan (OPZ) induced superoxide production. Zymosan particles were opsonized with human serum for 30 min at 37°C prior to use. The ratio of particles:cells was 20:1. In response to opsonized zymosan, cells deficient in p110␣ were reduced significantly in superoxide production compared with control cells (post-ANOVA Tukey test, P⬍0.01). Unstimulated cells were not significantly different from opsonized zymosan-stimulated, p110␣-deficient THP-1 cells (post-ANOVA Tukey test, P⬎0.05, n⫽7). (C) Nonopsonized zymosan did not induce superoxide production significantly in THP-1 cells (one-way ANOVA, P⫽0.3437). Cells were stimulated or not by zymosan particles at a ratio of 20:1, n ⫽ 3. Data are expressed in nanomoles superoxide (O2–) produced by 0.5 ⫻ 106 cells in 30 min. A P value ⬍0.05 was considered significant. Error bars indicate SEM.


1551

Fig. 3. Results of ELISA for TNF-␣, IL-10, IL-12p40, and IL-6 at 5 h and 18 h post-LPS stimulation of THP-1 cells, which transduced with control siRNA virus (shRNA control) or virus containing siRNA targeting p110␣ (p110␣ knockdown), were incubated or not with LPS at 100 ng/ml for 5 h or 18 h. Error bars indicate SD, n ⫽ 5. (A) LPS-induced TNF-␣ production was significantly higher in p110␣-deficient cells compared with controls at 5 h (post-ANOVA Tukey test, P⬍0.01) and 18 h (post-ANOVA Tukey test, P⬍0.05). (B) LPS-induced IL-10 production was significantly higher in p110␣-deficient cells compared with controls at 5 h (post-ANOVA Tukey test, P⬍0.001) and 18 h (post-ANOVA Tukey test, P⬍0.001). (C) LPS-induced IL-12p40 production was significantly lower in p110␣-deficient cells compared with controls at 5 h (post-ANOVA Tukey test, P⬍0.05) and 18 h (post-ANOVA Tukey test, P⬍0.001) and was not significantly different from unstimulated cells (P⬎0.05 for both time-points). (D) LPS-induced IL-6 production was significantly lower in p110␣-deficient cells compared with controls at 5 h (post-ANOVA Tukey test, P⬍0.05) and 18 h (post-ANOVA Tukey test, P⬍0.001) and was not significantly different from unstimulated cells (P⬎0.05 for both time-points).

1552



izing antibodies to TLR4 or TLR2 prior to LPS stimulation. At 18 h post-LPS exposure, IL-12p40 or IL-6 ELISAs were performed on the supernatants (Fig. 4). Preincubation with antiTLR4 antibody reduced IL-12 and IL-6 production by 54% (P⬍0.001) and 76% (P⬍0.01), respectively, relative to preincubation with anti-TLR2 antibody (Fig. 4, A and B). The quantities of LPS-induced IL-12 and IL-6 production in the presence of anti-TLR4 antibody were not significantly different from control cells incubated in the absence of antibody (⫹LPS, P⬎0.05). In contrast, preincubation with anti-TLR2 antibody only diminished zymosan-induced IL-12 (44%, P⬍0.01) but not IL-6 (P⬎0.05) production relative to preincubation with anti-TLR4 antibody, suggesting zymosan might trigger IL-6 production through other receptors as well. The amounts of IL-12 and IL-6 produced in response to zymosan in the presence of anti-TLR4 antibody were not significantly different from those produced by zymosan-treated control cells without neutralizing antibody preincubation (⫹Zy, P⬎0.05).

Neutralizing antibodies to TNF-␣ or IL-10 did not normalize the cytokine response of p110␣-deficient THP-1 cells Exogenous TNF-␣ has been reported to augment IL-10 production by monocytes [40]. We considered the possibility that enhanced TNF-␣ production by p110␣-deficient cells might bring about augmented IL-10 production via a paracrine effect. To examine this possibility, p110␣-deficient THP-1 cells were preincubated or not with anti-TNF-␣ or isotype-matched control antibodies. Figure 5 shows that anti-TNF-␣-neutralizing antibodies did not restore LPS-induced IL-10 (Fig. 5B), IL12p40 (Fig. 5C), or IL-6 (Fig. 5D) production by p110␣deficient cells to control levels. IL-10 has also been reported to suppress LPS-induced IL-6 and IL-12 production [41]. To investigate whether diminished LPS-induced IL-6 and IL-12 secretion by p110␣-deficient THP-1 cells was a result of enhanced IL-10 production, neutralizing antibody to IL-10 was added or not prior to stimulation of cell with LPS (Fig. 5, C and D). Preincubation with neutralizing anti-IL-10 antibody did not restore IL-6 or IL-12 production to levels seen in control cells. Finally, coincubation of cells with anti-TNF-␣- and anti-IL-10-neutralizing antibodies together also did not restore IL-6 or IL-12 to control levels (Fig. 6, A and B, respectively), ruling out a possible synergistic paracrine effect by enhanced TNF-␣ and IL-10 levels in the suppression of IL-12 and IL-6 production by p110␣deficient cells.

LPS-induced phosphorylation of p38 and JNK, but not ERK, is enhanced in p110␣-deficient THP-1 cells Activation of NF-␬B and p38 MAPK has been shown to be important in the regulation of LPS-induced cytokines such as TNF-␣, IL-6, and IL-12 [5, 6]. LPS is known to activate all three MAPKs including p38, JNK, and ERK [42]. As LPSstimulated, p110␣-deficient cells had an enhanced ability to secrete TNF-␣, but diminished capacity to secrete IL-6 and IL-12, we examined the phosphorylation status of MAPKs, as

Fig. 4. Effect of neutralizing antibodies to TLR4 and TLR2 on LPS and zymosan-induced IL-12p40 and IL-6 production in control, shRNA-transduced THP-1 cells. (A) IL-12p40 and (B) IL-6 ELISA were performed on supernatants derived from THP-1 cells, which had been transduced with control siRNA virus, preincubated for 5 h with TLR4 (aTLR4)- or TLR2 (aTLR2)-neutralizing antibody and then incubated with LPS (100 ng/ml) or serum-opsonized zymosan (Zy) for 18 h. Preincubation with anti-TLR4 antibody reduced IL-12p40 and IL-6 production by 54% (post-ANOVA Tukey test, P⬍0.001) and 76% (post-ANOVA Tukey test, P⬍0.01), respectively, relative to preincubation with anti-TLR2 antibody. In contrast, preincubation with anti-TLR2 antibody only diminished zymosan-induced IL-12p40 (44%, post-ANOVA Tukey test, P⬍0.01) but not IL-6 (post-ANOVA Tukey test, P⬎0.05) production. There was no significant difference between LPS stimulation alone versus anti-TLR2 with LPS or between zymosan alone and anti-TLR4 plus zymosan (P⬎0.05). Error bars indicate SD. The data are representative of two independent experiments.

well as that of NF-␬B p65 (Fig. 7). Of interest, antiphosphoAkt antibodies revealed that whereas LPS treatment resulted in increased phosphorylation of Akt in control cells, no significant increase in phospho-Akt was observed in p110␣-deficient cells (Fig. 7B). Of note, we found that phosphorylation of p38 and JNK (Fig. 7B) was enhanced significantly (both P⬍0.05, n⫽3) in p110␣-deficient cells at 20 and 30 min post-LPS stimulation compared with control cells. In contrast, phosphorylation of ERK was not altered significantly in p110␣-deficient cells when compared with controls and after normalization for protein loading (Fig. 7C). Conversely, the results shown in Figure 7D indicate that like p38 and JNK, phosphorylation of NF-␬B p65 was also clearly enhanced in p110␣-deficient THP-1 cells. Lee et al. PI-3K p110␣ gene silencing in monocytes

1553

Fig. 5. Effect of neutralizing anti-TNF-␣ or anti-IL-10 antibodies on LPS-induced IL-12p40 and IL-6 production by THP-1 cells, which when transduced with control siRNA virus (shRNA control, open bars) or virus containing siRNA targeting p110␣ (p110␣ knockdown, solid bars) were preincubated or not with anti-TNF-␣ antibody (aTNF␣), anti-IL-10 (aIL-10), or isotype-matched control antibody (rIgG or mIgG) at the indicated concentration for 5 h. Cells were then stimulated or not (No Tx) with LPS at 100 ng/ml for 5 h or 19 h. ELISAs for TNF-␣ (A), IL-10 (B), IL-12 (C), or IL-6 (D) were then performed on supernatants. (A and B) Samples treated with neutralizing antibodies TNF-␣ (A) or IL-10 (B) were all reduced significantly compared with control antibody plus LPS (post-ANOVA Tukey test, P⬍0.001). (C and D) shRNA control cells stimulated with LPS were significantly higher than all other treatment groups (post-ANOVA Tukey test, P⬍0.001), and the remaining groups were not significantly different from each other (post-ANOVA Tukey test, P⬎0.05). Error bars indicate SD. The data are representative of two independent experiments.

LPS-induced mRNA levels for TNF-␣, IL-6, IL-10, and IL-12p40

DISCUSSION

Using RT-PCR, we examined whether the changes observed in LPS-induced cytokine expression in p110␣-deficient cells could be explained by corresponding alterations in mRNA levels (Fig. 8). Consistent with reports that TNF-␣ is regulated transcriptionally and post-transcriptionally [43, 44], TNF-␣ mRNA levels from cells deficient in p110␣ were not enhanced significantly compared with controls, despite the fact that these cells produced more TNF-␣ (Fig. 3A). Likewise, increased secretion of IL-10 by p110␣-deficient cells was not reflected by increased IL-10 mRNA levels after LPS stimulation. In contrast, IL-12p40 mRNA levels in p110␣-deficient cells were significantly less than those of control cells, particularly at 4 h post-LPS stimulation. This was also found to be the case for IL-6, where mRNA levels were lower in p110␣-deficient cells when compared with control cells at 4 h and 16 h. These results suggest that diminished IL-12 and IL-6 production by p110␣deficient cells may in part have been related to events at the level of mRNA accumulation. In contrast, this does not appear to have been the case for enhanced TNF-␣ and IL-10 production.

Phagocytic cells play crucial roles in innate immunity and host defense by virtue of their abilities to recognize, ingest, and destroy invading microbes. PI-3K has been demonstrated to be an important regulator of Fc␥R and CR3-mediated phagocytosis [2, 23], but the specific isoform of PI-3K involved in mediating these events in human monocytic cells is not clear. Using microinjection of inhibitory antibodies, it was shown that p110␤ and to a lesser extent, p110␦, but not p110␣, was required for apoptotic cell and Fc␥R-mediated phagocytosis of IgG-opsonized RBC in murine macrophages [45]. These results are at variance with our findings in human monocytic cells, where we observed that phagocytosis of IgG-opsonized zymosan was p110␣ PI-3K-dependent (Fig. 1B). This discrepancy may be a result of species or cell-type differences. The differentiation state of the cell may also influence whether PI-3K regulates phagocytosis [46]. For example, it has been shown that undifferentiated THP-1 cells ingested IgG-opsonized particles, independent of PI-3K [47, 48], and Fc␥R-mediated phagocytosis by retinoic acid/IFN-␥-differentiated THP-1 cells was abrogated by PI-3K inhibitors [48].

1554



Fig. 6. Coincubation of neutralizing anti-TNF-␣ and anti-IL-10 antibodies did not restore LPS-induced IL-12p40 and IL-6 production. THP-1 cells transduced with control siRNA virus (shRNA control, open bars) or virus containing siRNA targeting p110␣ (p110␣ knockdown, solid bars) were preincubated or not with anti-TNF-␣ antibody, anti-IL-10 antibody, or isotype-matched control (ctl) antibodies (rIgG, mIgG, or IgG) at the indicated doses for 5 h. The cells were then stimulated or not (–) with LPS at 100 ng/ml for 19 h or 48 h. ELISAs for IL-12p40 (A) or IL-6 (B) were then performed on supernatants. Error bars indicate SD. The data are representative of two independent experiments.

Our data also indicate that phagocytosis of zymosan, but not serum-opsonized zymosan, is dependent on p110␣ (Fig. 1B). These results indicate that CR3-mediated phagocytosis does not require p110␣ PI-3K in our system. As CR3-mediated phagocytosis in murine macrophages has been shown to be attenuated by PI-3K inhibitors [23], our results implicate that another isoform of PI-3K is involved. Receptors, which can mediate phagocytosis of nonopsonized zymosan, include the MR and the ␤-glucan receptor [49 –51], and these could have contributed to the differential dependence on p110␣ PI-3K observed for uptake of nonopsonized versus opsonized zymosan [38, 52]. We investigated this possibility by mannose-blocking experiments and observed that phagocytosis of nonopsonized zymosan, but not serum-opsonized zymosan, was inhibited by soluble mannan (Fig. 1B). Thus, our results suggest that nonopsonized zymosan is internalized largely through the MR in a PI-3K p110␣-dependent manner. The NADPH-dependent oxidase is important in host defense and is also involved in the pathogenesis of inflammatory disorders such as arthritis [53], and ischemic/reperfusion injury [54]. Studies from Class IB p110␥ PI-3K KO mice suggested that fMLP-induced superoxide production was regulated by p110␥ [55, 56]. Although opsonzied zymosan-induced oxidase activation was normal in neutrophils from these p110␥ KO mice [55], experiments using PI-3K inhibitors showed that the oxidative burst in response to opsonized zymosan nevertheless required PI-3K [57]. Using THP-1 cells deficient in Class IA

p110␣ PI-3K, our results show that this isoform is required for the opsonized zymosan-induced oxidative burst in monocytic cells (Fig. 2B). This finding suggests that NADPH oxidase activation induced by a particulate agonist is brought about by a distinct PI-3K isoform from that induced by agonists, which act through G protein-coupled receptors (GPCRs), such as fMLP. In our system, nonopsonized zymosan did not induce significant superoxide production (Fig. 2C). However, nonopsonized zymosan has been reported to induce superoxide production in other contexts, such as in retinoic acid/vitamin D3-differentiated U-937 cells [58]. An apparent paradox was presented by the contrasting findings that phagocytosis of opsonized zymosan did not require p110␣ (Fig. 1B), and oxidase activation by opsonized zymosan did (Fig. 2C). These findings suggest that CR3 may not be the only receptor involved in phagocytosis of opsonized zymosan. A potential model is that opsonized zymosan-induced oxidase activation is mediated primarily by CR3 through its I/A domain interaction with iC3b coated on opsonized zymosan, and phagocytosis of opsonized zymosan may require other receptors, which can mediate p110␣-independent phagocytosis. Phorbol esters such as PMA mimic the action of the second messenger diacylglycerol (DAG) and can activate protein kinase C (PKC; classical and novel isoforms) and non-PKC phorbol ester/DAG receptors [59, 60]. As expected, we found that PMA triggered a strong oxidative burst in control cells, and cells deficient in p110␣ were effectively nonresponders Lee et al. PI-3K p110␣ gene silencing in monocytes

1555

Fig. 7. Analysis of kinase activation by Western blotting. Antibodies to (A) PI-3K p110␣, (B) phospho-Akt, phospho-JNK, and phospho-p38, (C) phospho-ERK, and (D) phospho-p65 NF-␬B were used to examine protein abundance or phosphorylation status in THP-1 cells, which were transduced with control siRNA virus (PI-3Kp110␣1) or virus containing siRNA targeting p110␣ (PI-3Kp110␣3). Cells were preincubated or not (–) with LPS at 100 ng/ml at various times (min). GAPDH was used as a protein-loading control. Results are representative of one of three independent experiments.

(Fig. 2A). These findings suggest that contradictory to the traditional view that PI-3K acts upstream of PKC [56], p110␣ is positioned downstream of PKC in activating the oxidase in mononuclear phagocytes. This finding is also consistent with the requirement for PI-3K metabolites for the recruitment of cytosolic oxidase components p47phox and p40phox to the mem-

Fig. 8. RT-PCR for TNF-␣, IL-12p40, IL-6, and IL-10 mRNA levels in LPS-stimulated THP-1 cells deficient in PI-3K p110␣. THP-1 cells transduced with control siRNA virus (␣1) or virus containing siRNA targeting p110␣ (␣3) were stimulated with 100 ng/ml LPS (⫹) or not (–) for the times indicated. Cells were then collected, and cDNAs were generated from the RNA preparations. Actin was used as loading control. Results are representative of three independent experiments.

1556


brane. Studies of trans-resveratrol (t-RVT), a naturally occurring polyphenolic compound found in grapes, are also consistent with a model in which p110␣ acts downstream of PKC in activating the oxidase [61]. It has been reported that t-RVT inhibited the PMA-induced oxidative burst in human monocytes, and an in vitro kinase screen revealed that t-RVT inhibited PI-3K by more than 40% and had no effect on 3-PI-dependent kinase 1 or Akt [61]. fMLP-stimulated U-937 cells also had increased PI-3K activity in antiphosphotyrosine immunoprecipitates, implicating Class IA PI-3K activation, and this activity was also inhibited by over 70% by t-RVT [61]. The ability of t-RVT to negatively modulate Class IA PI-3K activity coupled with its ability to inhibit PMA-induced oxidase activation [61] again suggests that the latter requires Class IA PI-3K activity. Thus, the findings we report reveal a novel role for p85/p110␣ PI-3K downstream of PKC, leading to activation of the monocyte oxidase. Although we believe the results presented in Figure 2A are most likely explained by PI-3K p110␣ acting downstream of PKC, alternative models could be proposed. For example, a yet-unidentified, non-PKC phorbol ester/DAG receptor may be involved upstream of PI-3K p110␣, leading to PMA-induced oxidase activation. The results of our studies focusing on activation of the oxidase are particularly informative when viewed in the context of studies that examined Class IB PI-3K p110␥⫺/⫺neutrophils [55], which from PI-3K p110␥ KO mice, had normal PMA- and opsonized zymosan-induced, oxidative burst responses but defective fMLP-induced burst activity when compared with control cells [55]. Although similar experiments in macrophages from these mice were not reported, these findings, together with http://www.jleukbio.org

our data, suggest that Class IA PI-3K p110␣ is required for oxidase activation downstream of CR3 and phorbol ester receptors, and Class IB PI-3K p110␥ is required for the GPCRstimulated oxidative burst. Regulation of cytokine production by PI-3K downstream of TLRs has been a subject of considerable interest [4]. The data presented herein suggest that p110␣ PI-3K positively regulates the production of the proinflammatory cytokines IL-12 and IL-6 and concurrently, negatively regulates the production of TNF-␣ and IL-10 in response to LPS (Fig. 3). It has been shown previously that the use of PI-3K inhibitors resulted in augmented production of TNF-␣ by LPS-treated human monocytic cells [5]. However, as LY294002 and wortmannin are neither PI-3K class- nor isoform-specific, it was not possible to assign this function to a specific PI-3K family member. The results of the present study show for the first time in human monocytic cells that an individual isoform of PI-3K—Class IA p110␣—is able to regulate the production of TNF-␣ and other cytokines. Previous studies in mice null for p85␣ PI-3K suggested that PI-3K negatively regulated LPS-induced IL-12 production by DC [6] and positively regulated TNF-␣ production in peritoneal washouts from mice challenged with colonic bacteria [62]. It is notable that these findings are the converse of those we describe here. These discrepancies may be explained by unexpected abnormalities in p85␣ KO mice. For example, DC from parental and p85 KO mice did not express detectable levels of p110␣ or p110␦ [6]. Conversely, mast cells from these animals expressed all three Class IA p110 isoforms, but the amount of p110␣ was reduced markedly [62]. Thus, it would appear to be important to interpret the results from cells derived from p85 KO animals within the context of their altered expression of each PI-3K Class IA catalytic isoform. A second anomaly in p85␣ null mice was the finding that they exhibited paradoxical increases in PI-3K activity, despite being null for p85␣ [63– 66]. Thus, the simplest way to reconcile the reports of enhanced IL-12 [6] and reduced TNF-␣ [62] production by cells from p85␣ KO mice with our converse findings in p110␣deficient cells is to conclude that these mice were not true Class IA PI-3K KOs. This conclusion is consistent with the observation that after LPS treatment of DC from p85␣ KO mice, Akt phosphorylation was only diminished mildly [67]. This indicates that significant Class IA PI-3K activity was still present in these p85␣ KO DC. To gain further insight into the role of PI-3K in regulating LPS-induced cytokine secretion, the effects of global PI-3K inhibitors, wortmannin and LY294002, on the secretion of cytokines were also examined for comparison with p110␣ knockdown cells. An important question to address was whether the phenotype we observed for p110␣ knockdowns would be reproduced by global inhibition of PI-3K. To address this, wild-type THP-1 cells were preincubated or not with wortmannin (50 nM) or LY294002 (10 ␮M) prior to LPS stimulation. With respect to each other, wortmannin and LY294002 had concordant and discordant effects on cytokine production. For example, wortmannin enhanced TNF-␣ secretion at 5 h and IL-12p40 at 18 h but inhibited IL-10 at 18 h (data not shown). Unlike wortmannin, LY294002 was null with respect to TNF-␣ production and attenuated secretion of IL-

12p40. Conversely, similar to wortmannin, LY294002 also inhibited IL-10 production at 18 h (data not shown). Discordant effects of LY294002 versus wortmannin have been reported previously [68] and may be a result of the fact that LY294002, at concentrations similar to those that inhibit PI3K, can also inhibit other kinases such as DNA protein kinase, mammalian target of rapamycin, and casein kinase 2, and wortmannin does not (reviewed in refs. [7, 69]). Thus, a recent report showing that wortmannin and LY294002 had opposite effects on NO production in LPS-stimulated macrophages [68] suggests that it may be incorrect to expect results with wortmannin and LY294002 to be consistently concordant. More germane to the present report, the results observed with neither LY294002 nor wortmannin were completely concordant with our data from p110␣ knockdown cells. Taken together, these findings lead to the conclusion that although these broadspectrum PI-3K inhibitors are useful for examining contributions by the PI-3K family of enzymes, they cannot contribute to an understanding of the precise roles (positive or negative) of individual isoforms. This is a result of their “specific” effects on multiple PI-3K isoforms as well as their “nonspecific” effects on other non-PI-3K targets (reviewed in refs. [7, 69]). Exogenous TNF-␣ has been reported to augment IL-10 production in monocytes [40], and IL-10 has also been demonstrated to suppress LPS-induced IL-6 and IL-12 production [41]. We considered the possibility that enhanced production of TNF-␣ by p110␣-deficient cells might have led to augmented IL-10 production through a paracrine effect and that this may have led to suppression of IL-6 and IL-12. Our data indicate that suppression of IL-12 and IL-6 was in fact not explained on this basis (Figs. 5 and 6). These experiments did not, however, rule out potential paracrine effects from other cytokines not assessed in this study. The independent effects of silencing p110␣ PI-3K on the several cytokines we studied suggest that this Class IA PI-3K cannot be classified as strictly proinflammatory or anti-inflammatory in regulating LPS-induced cytokine production. What appears clear, however, is that PI-3K p110␣ is required for IL-12 and IL-6 production, and it restricts the production of TNF-␣ and IL-10. As proinflammatory and anti-inflammatory signaling pathways have been reported to operate downstream of PI-3K (reviewed in ref. [4]), our results suggest that both of these effects may be regulated by p110␣ PI-3K, and PI-3K subunits can physically bind to key adaptors MyD88 [70] and translation initiation region (TIR) domain-containing adaptor-inducing IFN-␤ (TRIF) [71] of the TLR4 complex. These interactions support a model in which PI-3K can regulate signaling via MyD88-dependent or TRIF-dependent pathways. As one example, the transcription factor IFN regulatory factor 5 (IRF5) has been shown to regulate LPS-induced IL-6 and IL-12 production downstream of MyD88 [72], and endothelial cells expressing dominant-negative p85 had defective LPS-induced IL-6 production [30]. Taken together, these findings suggest the possibility that positive regulation of IRF5 by PI-3K p110␣ downstream of TLR4 could account for our observations of decreased, LPS-induced IL-6 and IL-12 production in THP-1 cells, where p110␣ had been silenced. A model incorporating the present findings about the potential roles of PI-3K p110␣ in LPS-induced cytokine expression Lee et al. PI-3K p110␣ gene silencing in monocytes

1557

Fig. 9. Proposed model for p110␣ PI-3K in regulating monocyte responses to LPS, and LPS and LPS-binding protein (LBP) bind to CD14 and present it to TLR4, where TLR4/myeloid differentiation protein 2 (MD2) undergoes dimerization and activation, which promotes tyrosine phosphorylation of MyD88 by an unknown kinase. This leads to the recruitment of p85/p110␣. TRIF has also been reported to be able to bind p110␣ directly, and recruitment increases with LPS stimulation [71]. IRF5 can bind directly to TNF receptor (TNFR)-associated factor 6 (TRAF6) and MyD88 and stimulate TNF-␣, IL-12p40, and IL-6 through binding to IFN-stimulated response element (ISRE) [72]. These effects are independent of p38 or JNK phosphorylation status. Conversely, IRF4 can mediate negative regulation of cytokine production via direct binding to MyD88, thereby competing with IRF5 binding [73]. In this proposed model, p110␣, via its effectors, stimulates IRF5, such that downstream IL-12 and IL-6 transcription are potentiated. TNF-␣ is also activated via NF-␬B, independent of IRF5, as IRF5⫺/⫺ mice still can secrete TNF-␣, albeit at a much lower level [72]. PI-3K p110␣ can limit, but not inhibit, TNF-␣ production. Activation of Akt by PI-3K may reduce p38 MAPK activation via inhibition of MAPK kinase kinase (MAPKKK), such as apoptosis signal-regulating kinase 1 (ASK-1) or MAPK/ERK kinase kinase 3 (MEKK-3) [74, 75]. This results in reduced phosphorylation of adenylate and uridylate-rich element-binding proteins (AREbp) by MAPK-activated protein kinase-2 (MK2), such that TNF-␣ mRNA is translationally repressed. TGF-␤-activated kinase 1 (TAK1) may be a target of PI-3K effectors, and inhibition of TAK1 may dampen NF-␬B activation and p38/JNK phosphorylation. IL-10 production may be regulated post-transcriptionally by p110␣, perhaps through a mechanism similar to TNF-␣. Dashed lines indicate how the complex of TRAF6 and the adaptor evolutionarily conserved signaling intermediate in Toll pathways (ECSIT) interacts specifically with MEKK-1 (another MAPKKK) and appears to function in this pathway by facilitating the processing of MEKK-1 [76]. Arrows indicate activation, and blunt arrows indicate inhibition. SHIP, SH2 domain-containing inositol phosphatase; PTEN, phosphatase and tensin homolog deleted from chromosome 10; IRAK, IL-1R-associated kinase; Mal, MyD88adaptor-like protein; TRAM, Toll-IL-1R domain-containing adaptor inducing IFN-␤-related adaptor molecule; TAB, TAK-binding protein; TANK, TNFRassociated factor family member-associated NF-␬B activator; IKK, I␬B kinase; TBK1, TANK-binding kinase 1; Elk1, E-26-like protein 1; PRD, positive regulatory domain; Sp1/3, . See text for more details. Adapted from refs. [4, 28, 70, 71, 76 –79].

in human monocytic cells is shown in Figure 9. We propose that p110␣ may function to activate IRF5, similar to the mechanism described for TLR3/IRF3 signaling in HEK293 cells in which IRF3 was activated through phosphorylation by the PI-3K effector Akt [77]. The overall effect of IRF5 activation is the induction of proinflammatory cytokines such IL12p40 and IL-6. This would be consistent with diminished IL-12p40 and IL-6 cytokine mRNA levels seen in LPS-stimulated, p110␣-deficient THP-1 cells. Thus far, the precise mechanism of IRF5 activation is not understood [72], but our findings suggest one potential candidate may be PI-3K p110␣. TNF-␣ may also be induced positively by IRF5, but its transcription in THP-1 cells may be less dependent on IRF5 than on NF-␬B, as suggested by the fact that IRF5⫺/⫺ mice were still able to secrete TNF-␣, albeit at a reduced level [72]. In the working model, we propose, therefore, despite reduced IRF5 activation, increased NF-␬B and p38 phosphorylation in LPS1558


stimulated, p110␣-deficient cells (Fig. 7) may overcompensate and lead to enhanced TNF-␣ production. A mechanism by which p110␣ PI-3K may normally negatively regulate TNF-␣ production might involve Akt phosphorylation and inhibition of a MAPKKK such as ASK-1 or MEKK-3 [74, 75]. This could lead to restricted activation of p38, less MK2 activity, and correspondingly diminished phosphorylation of AREbp. As a result, these proteins would retain their ability to inhibit translation (Fig. 9). Limited activation of p38 MAPK could also restrict IL-10 production to some extent by restraining activation of Sp1 [80]. It is interesting that in contrast to phospho-p38 MAPK and phospho-JNK, phospho-ERK was not elevated significantly in LPS-stimulated, p110␣-deficient cells (Fig. 7). This suggests that p110␣ may regulate p38 MAPK and JNK selectively. How this may come about remains to be determined. PI-3K effectors such as Akt may inhibit TAK1, leading to reduced activation of http://www.jleukbio.org

p38, JNK, and NF-␬B, as human monocytic cells expressing kinase-dead TAK1 showed reduced IKK␤ activity in response to LPS treatment, as well as a reduction in p38 and JNK phosphorylation, but not that of ERK [81]. In summary, using lentiviral-transduced THP-1 cells deficient in PI-3K p110␣, we report prominent and selective roles for p110␣ in regulating phagocytosis, activation of the phagocyte oxidase, and in LPS-induced cytokine production. The latter findings in particular highlight the complexity of the roles of PI-3K in regulating diverse responses downstream of TLR4. Our data show that p110␣ can regulate several important cytokines independently of other PI-3K isoforms. This observation may have been masked by other approaches such as those that used global PI-3K inhibitors or mice with targeted disruptions of PI-3K genes.

ACKNOWLEDGMENTS This work was supported by Canadian Institutes of Health Research (CIHR) grant MOP-8633 (N. E. R.) and National Institutes of Health grant R01 AI34879 (W. M. N.). J. S. L. was supported by a M.D./Ph.D. Studentship from CIHR and a Research Doctoral Trainee Award from Michael Smith Foundation for Health Research.

REFERENCES 1. Seljelid, R., Eskeland, T. (1993) The biology of macrophages: I General principles and properties. Eur. J. Haematol. 51, 267–275. 2. Araki, N., Johnson, M. T., Swanson, J. A. (1996) A role for phosphoinositide 3-kinase in the completion of macropinocytosis and phagocytosis by macrophages. J. Cell Biol. 135, 1249 –1260. 3. Yamamori, T., Inanami, O., Nagahata, H., Kuwabara, M. (2004) Phosphoinositide 3-kinase regulates the phosphorylation of NADPH oxidase component p47(phox) by controlling cPKC/PKC␦ but not Akt. Biochem. Biophys. Res. Commun. 316, 720 –730. 4. Fukao, T., Koyasu, S. (2003) PI3K and negative regulation of TLR signaling. Trends Immunol. 24, 358 –363. 5. Guha, M., Mackman, N. (2002) The phosphatidylinositol 3-kinase-Akt pathway limits lipopolysaccharide activation of signaling pathways and expression of inflammatory mediators in human monocytic cells. J. Biol. Chem. 277, 32124 –32132. 6. Fukao, T., Tanabe, M., Terauchi, Y., Ota, T., Matsuda, S., Asano, T., Kadowaki, T., Takeuchi, T., Koyasu, S. (2002) PI3K-mediated negative feedback regulation of IL-12 production in DCs. Nat. Immunol. 3, 875– 881. 7. Vanhaesebroeck, B., Leevers, S. J., Ahmadi, K., Timms, J., Katso, R., Driscoll, P. C., Woscholski, R., Parker, P. J., Waterfield, M. D. (2001) Synthesis and function of 3-phosphorylated inositol lipids. Annu. Rev. Biochem. 70, 535– 602. 8. Herrera-Velit, P., Knutson, K. L., Reiner, N. E. (1997) Phosphatidylinositol 3-kinase-dependent activation of protein kinase C-␨ in bacterial lipopolysaccharide-treated human monocytes. J. Biol. Chem. 272, 16445–16452. 9. Hmama, Z., Nandan, D., Sly, L., Knutson, K. L., Herrera-Velit, P., Reiner, N. E. (1999) 1␣,25-Dihydroxyvitamin D(3)-induced myeloid cell differentiation is regulated by a vitamin D receptor-phosphatidylinositol 3-kinase signaling complex. J. Exp. Med. 190, 1583–1594. 10. Petiot, A., Ogier-Denis, E., Blommaart, E. F., Meijer, A. J., Codogno, P. (2000) Distinct classes of phosphatidylinositol 3⬘-kinases are involved in signaling pathways that control macroautophagy in HT-29 cells. J. Biol. Chem. 275, 992–998 [Erratum in: J. Biol. Chem., 2000, 275, 12360]. 11. Vanhaesebroeck, B., Jones, G. E., Allen, W. E., Zicha, D., HooshmandRad, R., Sawyer, C., Wells, C., Waterfield, M. D., Ridley, A. J. (1999) Distinct PI(3)Ks mediate mitogenic signaling and cell migration in macrophages. Nat. Cell Biol. 1, 69 –71.

12. Siddhanta, U., McIlroy, J., Shah, A., Zhang, Y., Backer, J. M. (1998) Distinct roles for the p110␣ and hVPS34 phosphatidylinositol 3⬘-kinases in vesicular trafficking, regulation of the actin cytoskeleton, and mitogenesis. J. Cell Biol. 143, 1647–1659. 13. Vieira, O. V., Botelho, R. J., Rameh, L., Brachmann, S. M., Matsuo, T., Davidson, H. W., Schreiber, A., Backer, J. M., Cantley, L. C., Grinstein, S. (2001) Distinct roles of class I and class III phosphatidylinositol 3-kinases in phagosome formation and maturation. J. Cell Biol. 155, 19 –25. 14. Stephens, L. R., Hughes, K. T., Irvine, R. F. (1991) Pathway of phosphatidylinositol(3,4,5)-trisphosphate synthesis in activated neutrophils. Nature 351, 33–39. 15. Hawkins, P. T., Jackson, T. R., Stephens, L. R. (1992) Platelet-derived growth factor stimulates synthesis of PtdIns(3,4,5)P3 by activating a PtdIns(4,5)P2 3-OH kinase. Nature 358, 157–159. 16. Shepherd, P. R., Nave, B. T., Rincon, J., Nolte, L. A., Bevan, A. P., Siddle, K., Zierath, J. R., Wallberg-Henriksson, H. (1997) Differential regulation of phosphoinositide 3-kinase adapter subunit variants by insulin in human skeletal muscle. J. Biol. Chem. 272, 19000 –19007. 17. Janssen, J. W., Schleithoff, L., Bartram, C. R., Schulz, A. S. (1998) An oncogenic fusion product of the phosphatidylinositol 3-kinase p85␤ subunit and HUMORF8, a putative deubiquitinating enzyme. Oncogene 16, 1767–1772. 18. Dey, B. R., Furlanetto, R. W., Nissley, S. P. (1998) Cloning of human p55 ␥, a regulatory subunit of phosphatidylinositol 3-kinase, by a yeast twohybrid library screen with the insulin-like growth factor-I receptor. Gene 209, 175–183. 19. Cox, D., Tseng, C. C., Bjekic, G., Greenberg, S. (1999) A requirement for phosphatidylinositol 3-kinase in pseudopod extension. J. Biol. Chem. 274, 1240 –1247. 20. Ninomiya, N., Hazeki, K., Fukui, Y., Seya, T., Okada, T., Hazeki, O., Ui, M. (1994) Involvement of phosphatidylinositol 3-kinase in Fc ␥ receptor signaling. J. Biol. Chem. 269, 22732–22737. 21. Strzelecka, A., Pyrzynska, B., Kwiatkowska, K., Sobota, A. (1997) Syk kinase, tyrosine-phosphorylated proteins and actin filaments accumulate at forming phagosomes during Fc␥ receptor-mediated phagocytosis. Cell Motil. Cytoskeleton 38, 287–296. 22. Crowley, M. T., Costello, P. S., Fitzer-Attas, C. J., Turner, M., Meng, F. Y., Lowell, C., Tybulewicz, V. L. J., DeFranco, A. L. (1997) A critical role for Syk in signal transduction and phagocytosis mediated by Fc␥ receptors on macrophages. J. Exp. Med. 186, 1027–1039. 23. Cox, D., Dale, B. M., Kashiwada, M., Helgason, C. D., Greenberg, S. (2001) A regulatory role for Src homology 2 domain-containing inositol 5⬘-phosphatase (SHIP) in phagocytosis mediated by Fc ␥ receptors and complement receptor 3 (␣(M)␤(2); CD11b/CD18). J. Exp. Med. 193, 61–71. 24. Noack, D., Rae, J., Cross, A. R., Ellis, B. A., Newburger, P. E., Curnutte, J. T., Heyworth, P. G. (2001) Autosomal recessive chronic granulomatous disease caused by defects in NCF-1, the gene encoding the phagocyte p47-phox: mutations not arising in the NCF-1 pseudogenes. Blood 97, 305–311. 25. Nauseef, W. M. (2004) Assembly of the phagocyte NADPH oxidase. Histochem. Cell Biol. 122, 277–291. 26. Simonsen, A., Stenmark, H. (2001) PX domains: attracted by phosphoinositides. Nat. Cell Biol. 3, E179 –E182. 27. Takeda, K., Kaisho, T., Akira, S. (2003) Toll-like receptors. Annu. Rev. Immunol. 21, 335–376. 28. Hebeis, B. J., Vigorito, E., Turner, M. (2004) The p110␦ subunit of phosphoinositide 3-kinase is required for the lipopolysaccharide response of mouse B cells. Biochem. Soc. Trans. 32, 789 –791. 29. Okugawa, S., Ota, Y., Kitazawa, T., Nakayama, K., Yanagimoto, S., Tsukada, K., Kawada, M., Kimura, S. (2003) Janus kinase 2 is involved in lipopolysaccharide-induced activation of macrophages. Am. J. Physiol. Cell Physiol. 285, C399 –C408. 30. Li, X., Tupper, J. C., Bannerman, D. D., Winn, R. K., Rhodes, C. J., Harlan, J. M. (2003) Phosphoinositide 3 kinase mediates Toll-like receptor 4-induced activation of NF-␬ B in endothelial cells. Infect. Immun. 71, 4414 – 4420. 31. Sizemore, N., Leung, S., Stark, G. R. (1999) Activation of phosphatidylinositol 3-kinase in response to interleukin-1 leads to phosphorylation and activation of the NF-␬B p65/RelA subunit. Mol. Cell. Biol. 19, 4798 – 4805. 32. Ishii, K. J., Takeshita, F., Gursel, I., Gursel, M., Conover, J., Nussenzweig, A., Klinman, D. M. (2002) Potential role of phosphatidylinositol 3 kinase, rather than DNA-dependent protein kinase, in CpG DNA-induced immune activation. J. Exp. Med. 196, 269 –274. 33. Jones, B. W., Heldwein, K. A., Means, T. K., Saukkonen, J. J., Fenton, M. J. (2001) Differential roles of Toll-like receptors in the elicitation of


1559

34.

35.

36.

37.

38.

39. 40.

41.

42. 43.

44.

45.

46. 47.

48.

49.

50.

51.

52.

53.

54.

55.

proinflammatory responses by macrophages. Ann. Rheum. Dis. 60 (Suppl. 3), iii6 –12. Weinstein, S. L., Finn, A. J., Dave, S. H., Meng, F., Lowell, C. A., Sanghera, J. S., DeFranco, A. L. (2000) Phosphatidylinositol 3-kinase and mTOR mediate lipopolysaccharide-stimulated nitric oxide production in macrophages via interferon-␤. J. Leukoc. Biol. 67, 405– 414. Lee, J. S., Hmama, Z., Mui, A., Reiner, N. E. (2004) Stable gene silencing in human monocytic cell lines using lentiviral-delivered small interference RNA. Silencing of the p110␣ isoform of phosphoinositide 3-kinase reveals differential regulation of adherence induced by 1␣,25-dihydroxycholecalciferol and bacterial lipopolysaccharide. J. Biol. Chem. 279, 9379 –9388. DeLeo, F. R., Jutila, M. A., Quinn, M. T. (1996) Characterization of peptide diffusion into electropermeabilized neutrophils. J. Immunol. Methods 198, 35– 49. DeLeo, F. R., Allen, L. A., Apicella, M., Nauseef, W. M. (1999) NADPH oxidase activation and assembly during phagocytosis. J. Immunol. 163, 6732– 6740. Suzuki, T., Ohno, N., Ohshima, Y., Yadomae, T. (1998) Soluble mannan and ␤-glucan inhibit the uptake of Malassezia furfur by human monocytic cell line, THP-1. FEMS Immunol. Med. Microbiol. 21, 223–230. Underhill, D. M., Ozinsky, A. (2002) Phagocytosis of microbes: complexity in action. Annu. Rev. Immunol. 20, 825– 852. Giambartolomei, G. H., Dennis, V. A., Lasater, B. L., Murthy, P. K., Philipp, M. T. (2002) Autocrine and exocrine regulation of interleukin-10 production in THP-1 cells stimulated with Borrelia burgdorferi lipoproteins. Infect. Immun. 70, 1881–1888. Murthy, P. K., Dennis, V. A., Lasater, B. L., Philipp, M. T. (2000) Interleukin-10 modulates proinflammatory cytokines in the human monocytic cell line THP-1 stimulated with Borrelia burgdorferi lipoproteins. Infect. Immun. 68, 6663– 6669. Rao, K. M. (2001) MAP kinase activation in macrophages. J. Leukoc. Biol. 69, 3–10. Kotlyarov, A., Neininger, A., Schubert, C., Eckert, R., Birchmeier, C., Volk, H. D., Gaestel, M. (1999) MAPKAP kinase 2 is essential for LPS-induced TNF-␣ biosynthesis. Nat. Cell Biol. 1, 94 –97. Han, J., Brown, T., Beutler, B. (1990) Endotoxin-responsive sequences control cachectin/tumor necrosis factor biosynthesis at the translational level. J. Exp. Med. 171, 465– 475. Leverrier, Y., Okkenhaug, K., Sawyer, C., Bilancio, A., Vanhaesebroeck, B., Ridley, A. J. (2003) Class I phosphoinositide 3-kinase p110␤ is required for apoptotic cell and Fc␥ receptor-mediated phagocytosis by macrophages. J. Biol. Chem. 278, 38437–38442. Garcia-Garcia, E., Rosales, C. (2002) Signal transduction during Fc receptor-mediated phagocytosis. J. Leukoc. Biol. 72, 1092–1108. Garcia-Garcia, E., Sanchez-Mejorada, G., Rosales, C. (2001) Phosphatidylinositol 3-kinase and ERK are required for NF-␬B activation but not for phagocytosis. J. Leukoc. Biol. 70, 649 – 658. Garcia-Garcia, E., Rosales, R., Rosales, C. (2002) Phosphatidylinositol 3-kinase and extracellular signal-regulated kinase are recruited for Fc receptor-mediated phagocytosis during monocyte-to-macrophage differentiation. J. Leukoc. Biol. 72, 107–114. Ross, G. D., Cain, J. A., Lachmann, P. J. (1985) Membrane complement receptor type three (CR3) has lectin-like properties analogous to bovine conglutinin as functions as a receptor for zymosan and rabbit erythrocytes as well as a receptor for iC3b. J. Immunol. 134, 3307–3315. Speert, D. P., Silverstein, S. C. (1985) Phagocytosis of unopsonized zymosan by human monocyte-derived macrophages: maturation and inhibition by mannan. J. Leukoc. Biol. 38, 655– 658. Giaimis, J., Lombard, Y., Fonteneau, P., Muller, C. D., Levy, R., MakayaKumba, M., Lazdins, J., Poindron, P. (1993) Both mannose and ␤-glucan receptors are involved in phagocytosis of unopsonized, heat-killed Saccharomyces cerevisiae by murine macrophages. J. Leukoc. Biol. 54, 564 – 571. Diaz-Silvestre, H., Espinosa-Cueto, P., Sanchez-Gonzalez, A., EsparzaCeron, M. A., Pereira-Suarez, A. L., Bernal-Fernandez, G., Espitia, C., Mancilla, R. (2005) The 19-kDa antigen of Mycobacterium tuberculosis is a major adhesin that binds the mannose receptor of THP-1 monocytic cells and promotes phagocytosis of mycobacteria. Microb. Pathog. 39, 97–107. Hultqvist, M., Holmdahl, R. (2005) Ncf1 (p47phox) polymorphism determines oxidative burst and the severity of arthritis in rats and mice. Cell. Immunol. 233, 97–101. Korthuis, R. J., Granger, D. N. (1993) Reactive oxygen metabolites, neutrophils, and the pathogenesis of ischemic-tissue/reperfusion. Clin. Cardiol. 16, I19 –I26. Hirsch, E., Katanaev, V. L., Garlanda, C., Azzolino, O., Pirola, L., Silengo, L., Sozzani, S., Mantovani, A., Altruda, F., Wymann, M. P. (2000) Central

1560


56. 57.

58.

59. 60. 61. 62.

63.

64.

65.

66.

67. 68.

69. 70.

71.

72.

73. 74. 75.

76.

role for G protein-coupled phosphoinositide 3-kinase ␥ in inflammation. Science 287, 1049 –1053. Wymann, M. P., Sozzani, S., Altruda, F., Mantovani, A., Hirsch, E. (2000) Lipids on the move: phosphoinositide 3-kinases in leukocyte function. Immunol. Today 21, 260 –264. Yamamori, T., Inanami, O., Sumimoto, H., Akasaki, T., Nagahata, H., Kuwabara, M. (2002) Relationship between p38 mitogen-activated protein kinase and small GTPase Rac for the activation of NADPH oxidase in bovine neutrophils. Biochem. Biophys. Res. Commun. 293, 1571–1578. Le Cabec, V., Cols, C., Maridonneau-Parini, I. (2000) Nonopsonic phagocytosis of zymosan and Mycobacterium kansasii by CR3 (CD11b/CD18) involves distinct molecular determinants and is or is not coupled with NADPH oxidase activation. Infect. Immun. 68, 4736 – 4745. Caloca, M. J., Wang, H., Kazanietz, M. G. (2003) Characterization of the Rac-GAP (Rac-GTPase-activating protein) activity of ␤2-chimaerin, a ‘non-protein kinase C’ phorbol ester receptor. Biochem. J. 375, 313–321. Yang, C., Kazanietz, M. G. (2003) Divergence and complexities in DAG signaling: looking beyond PKC. Trends Pharmacol. Sci. 24, 602– 608. Poolman, T. M., Ng, L. L., Farmer, P. B., Manson, M. M. (2005) Inhibition of the respiratory burst by resveratrol in human monocytes: correlation with inhibition of PI3K signaling. Free Radic. Biol. Med. 39, 118 –132. Fukao, T., Yamada, T., Tanabe, M., Terauchi, Y., Ota, T., Takayama, T., Asano, T., Takeuchi, T., Kadowaki, T., Hata, Ji, J., Koyasum, S. (2002) Selective loss of gastrointestinal mast cells and impaired immunity in PI3K-deficient mice. Nat. Immunol. 3, 295–304. Terauchi, Y., Tsuji, Y., Satoh, S., Minoura, H., Murakami, K., Okuno, A., Inukai, K., Asano, T., Kaburagi, Y., Ueki, K., et al. (1999) Increased insulin sensitivity and hypoglycaemia in mice lacking the p85 ␣ subunit of phosphoinositide 3-kinase. Nat. Genet. 21, 230 –235. Fruman, D. A., Mauvais-Jarvis, F., Pollard, D. A., Yballe, C. M., Brazil, D., Bronson, R. T., Kahn, C. R., Cantley, L. C. (2000) Hypoglycaemia, liver necrosis and perinatal death in mice lacking all isoforms of phosphoinositide 3-kinase p85 ␣. Nat. Genet. 26, 379 –382. Mauvais-Jarvis, F., Ueki, K., Fruman, D. A., Hirshman, M. F., Sakamoto, K., Goodyear, L. J., Iannacone, M., Accili, D., Cantley, L. C., Kahn, C. R. (2002) Reduced expression of the murine p85␣ subunit of phosphoinositide 3-kinase improves insulin signaling and ameliorates diabetes. J. Clin. Invest. 109, 141–149. Ueki, K., Yballe, C. M., Brachmann, S. M., Vicent, D., Watt, J. M., Kahn, C. R., Cantley, L. C. (2002) Increased insulin sensitivity in mice lacking p85␤ subunit of phosphoinositide 3-kinase. Proc. Natl. Acad. Sci. USA 99, 419 – 424. Vanhaesebroeck, B., Ali, K., Bilancio, A., Geering, B., Foukas, L. C. (2005) Signaling by PI3K isoforms: insights from gene-targeted mice. Trends Biochem. Sci. 30, 194 –204. Hazeki, K., Kinoshita, S., Matsumura, T., Nigorikawa, K., Kubo, H., Hazeki, O. (2006) Opposite effects of wortmannin and 2-(4-morpholinyl)8-phenyl-1(4H)-benzopyran-4-one hydrochloride on Toll-like receptormediated nitric oxide production: negative regulation of nuclear factor{␬}B by phosphoinositide 3-kinase. Mol. Pharmacol. 69, 1717–1724. Wetzker, R., Rommel, C. (2004) Phosphoinositide 3-kinases as targets for therapeutic intervention. Curr. Pharm. Des. 10, 1915–1922. Ojaniemi, M., Glumoff, V., Harju, K., Liljeroos, M., Vuori, K., Hallman, M. (2003) Phosphatidylinositol 3-kinase is involved in Toll-like receptor 4-mediated cytokine expression in mouse macrophages. Eur. J. Immunol. 33, 597– 605. Aksoy, E., Vanden, B. W., Detienne, S., Amraoui, Z., Fitzgerald, K. A., Haegeman, G., Goldman, M., Willems, F. (2005) Inhibition of phosphoinositide 3-kinase enhances TRIF-dependent NF-␬ B activation and IFN-␤ synthesis downstream of Toll-like receptor 3 and 4. Eur. J. Immunol. 35, 2200 –2209. Takaoka, A., Yanai, H., Kondo, S., Duncan, G., Negishi, H., Mizutani, T., Kano, S., Honda, K., Ohba, Y., Mak, T. W., Taniguchi, T. (2005) Integral role of IRF-5 in the gene induction programme activated by Toll-like receptors. Nature 434, 243–249. Sarkar, S. N., Peters, K. L., Elco, C. P., Sakamoto, S., Pal, S., Sen, G. C. (2004) Novel roles of TLR3 tyrosine phosphorylation and PI3 kinase in double-stranded RNA signaling. Nat. Struct. Mol. Biol. 11, 1060 –1067. Kim, A. H., Khursigara, G., Sun, X., Franke, T. F., Chao, M. V. (2001) Akt phosphorylates and negatively regulates apoptosis signal-regulating kinase 1. Mol. Cell Biol. 21, 893–901. Gratton, J. P., Morales-Ruiz, M., Kureishi, Y., Fulton, D., Walsh, K., Sessa, W. C. (2001) Akt down-regulation of p38 signaling provides a novel mechanism of vascular endothelial growth factor-mediated cytoprotection in endothelial cells. J. Biol. Chem. 276, 30359 –30365. Ma, W., Lim, W., Gee, K., Aucoin, S., Nandan, D., Kozlowski, M., az-Mitoma, F., Kumar, A. (2001) The p38 mitogen-activated kinase pathway regulates the human interleukin-10 promoter via the activation of Sp1


transcription factor in lipopolysaccharide-stimulated human macrophages. J. Biol. Chem. 276, 13664 –13674. 77. Lee, J., Mira-Arbibe, L., Ulevitch, R. J. (2000) TAK1 regulates multiple protein kinase cascades activated by bacterial lipopolysaccharide. J. Leukoc. Biol. 68, 909 –915. 78. Negishi, H., Ohba, Y., Yanai, H., Takaoka, A., Honma, K., Yui, K., Matsuyama, T., Taniguchi, T., Honda, K. (2005) Negative regulation of Toll-like-receptor signaling by IRF-4. Proc. Natl. Acad. Sci. U.S.A. 102, 15989 –15994.

79. Kopp, E., Medzhitov, R., Carothers, J., Xiao, C., Douglas, I., Janeway, C. A., Ghosh, S. (1999) ECSIT is an evolutionarily conserved intermediate in the Toll/IL-1 signal transduction pathway. Genes Dev. 13, 2059 –2071. 80. Sly, L. M., Rauh, M. J., Kalesnikoff, J., Buchse, T., Krystal, G. (2003) SHIP, SHIP2, and PTEN activities are regulated in vivo by modulation of their protein levels: SHIP in up-regulated in macrophages and mast cells by lipopolysaccharide. Exp. Hematol. 31, 1170 –1181. 81. Akira, S., and Takeda, K. (2004) Toll-like receptor signalling. Nat. Rev. Immunol. 4, 499 –511.


1561