Phagocytosis of microparticles increases responsiveness of macrophage-like cell lines U937 and THP-1 to bacterial lipopolysaccharide and lipopeptide

Following bacterial infection, macrophages produce pro-inflammatory cytokines in response to bacterial cell components, including lipopolysaccharide (LPS) and lipopeptide, and simultaneously phagocytize and digest the invading bacteria. To study the effects of phagocytosis on pro-inflammatory responses, we determined if phagocytosis of polystyrene latex beads with ~ 1 µm diameter increases pro-inflammatory cytokine expression by human macrophage-like U937 and THP-1 cells stimulated with LPS. Treating macrophage-like cells with beads coated with IgG to facilitate Fcγ receptor-mediated phagocytosis increased LPS-induced expression of pro-inflammatory cytokines, including tumor necrosis factor-alpha, interleukin-1 beta, and interleukin-6. Treatment with beads coated with poly-l-lysine to facilitate Fcγ receptor–independent phagocytosis also increased LPS-induced cytokine expression. Our results indicate that LPS-induced pro-inflammatory responses are enhanced by bead phagocytosis regardless of the uptake mechanism. Additionally, phagocytosis enhanced LPS-induced NF-κB activation, suggesting that Toll-like receptor (TLR) 4 signaling is enhanced by phagocytosis. Furthermore, bead phagocytosis enhanced pro-inflammatory responses in U937 cells stimulated with lipopeptide, a ligand for the TLR2/TLR6 heterodimeric receptor. In conclusion, microparticle phagocytosis by macrophage-like U937 and THP-1 cells enhances the innate immune response induced by bacterial components.

. Phagocytosis of IgG-coated beads increases LPS-induced TNFα expression from U937 cells. Beads were added to U937 cells at a bead-to-cell ratio of 500. The number of phagocytized beads (a-c) or the amount of TNFα in culture supernatant (d-g) was determined. Daggers ( †) indicate values below the detection limit. The data shown are mean values with standard deviations derived from triplicate measurements. Data were analyzed by the t-test. *P < 0.05. (a) Cells were cultured with beads pre-treated with the indicated concentrations of human IgG for 4 h. (b) Cells were cultured with indicated concentrations of Cytochalasin D for 1 h. Next, the beads pre-treated with 1000 µg/ml IgG were added to the culture and cultured for an additional 4 h. (c) Cells were cultured with the indicated concentrations of human IgG for 4 h. The beads pre-treated with 1000 µg/ml IgG were added to the culture and the cells were cultured for an additional 4 h. (d) Cells stimulated with 100 ng/ml LPS were cultured simultaneously either with ( +) or without (−) beads pre-treated with the indicated concentrations of human IgG. (e) Cells stimulated with indicated concentrations of LPS were cultured simultaneously with (gray bar) or without (white bar) beads pre-treated with 1000 µg/ml human IgG. (f) Cells stimulated with 10 ng/ml LPS were cultured simultaneously with (closed circle, Beads +) or without (open circle, Beads −) beads pre-treated with 1000 µg/ml human IgG for 0, 1,2,4,8,16, and 24 h. (g) Cells were cultured with indicated concentrations of human IgG for 4 h prior to addition of LPS and beads. Next, cells stimulated with 10 ng/ml LPS were cultured simultaneously either with (gray bar) or without (white bar) beads pre-treated with 1000 µg/ml human IgG.

Fcγ receptor-independent phagocytosis also enhances LPS-induced pro-inflammatory responses.
To further investigate the effects of phagocytosis on cytokine expression, polystyrene latex beads with an approximate diameter of 1 µm were covalently coupled to poly-L-lysine (PLL) before being used for this study. Positively charged particles, including PLL-coated beads, are known to be efficiently phagocytized by macrophages independently of the Fcγ receptor 20 . Phagocytosis of PLL-coated beads was observed by THP-1 cells; we found that phagocytosis of PLL-coated beads by THP-1 cells (Fig. 4a) was much greater than phagocytosis of uncoated beads (Fig. 3a). In contrast, PLL-dependent phagocytosis was not observed in U937 cells (data not shown). We therefore used THP-1 cells for further analysis. Pre-treating THP-1 cells with 100 µg/ml IgG inhibited phagocytosis of IgG-coated beads by 42% but not PLL-coated beads (Fig. 4a). In contrast, pre-treatment of THP-1 cells with 10 µM cytochalasin D inhibited Figure 2. Phagocytosis of IgG-coated beads increases LPS-induced expression of IL-1β and IL-6 by U937 cells. U937 cells stimulated with indicated concentrations of LPS were cultured simultaneously either with (gray bar) or without (white bar) beads pre-treated with 1000 µg/ml human IgG at a ratio of 500 beads/cell. After cultivation, the amounts of IL-1β (a) and IL-6 (b) were determined. The data shown are mean values with standard deviations derived from triplicate measurements. Daggers ( †) indicate below the detection limit. Data were analyzed by the t-test. *P < 0.05. www.nature.com/scientificreports/ phagocytosis of PLL-coated beads by 65% (Fig. 4b).

Scientific Reports
These results indicate that phagocytosis of PLL-coated beads is not mediated by the Fcγ receptor. Thus, we determined whether phagocytosis of PLL-coated beads affects cytokine expression by LPS-stimulated THP-1 cells. Phagocytosis of PLL-coated beads increased TNFα, IL-1β, and IL-6 expression in cells stimulated with 10 ng/ml LPS by 1.7-fold, 8.0-fold, and 3.1-fold, respectively (Fig. 5). Phagocytosis of PLL-coated beads alone slightly increased the cytokine expression, but the increments were much lower than those with LPS-stimulated cells (Fig. 5). These results indicate that Fcγ receptor-independent phagocytosis also enhances LPS-induced pro-inflammatory responses in THP-1 cells.  www.nature.com/scientificreports/ Phagocytosis enhances LPS-induced activation of NF-κB. We found that bead phagocytosis enhances the expression of pro-inflammatory cytokines from macrophage-like U937 and THP-1 cells stimulated with LPS. In order to determine whether enhancement occurred at the level of transcription or post-transcriptionally, we examined the effects of phagocytosis of beads on NF-κB activation in U937 cells. LPS-induced activation of NF-κB, which was represented by two mobility-shifted bands 21 , was observed 1 h after stimulation, reached a maximum level at 2 h after stimulation, and decreased at 4 h after stimulation (Fig. 6a). Phagocytosis of IgG-coated beads increased NF-κB activation at all of these time points, and NF-κB activation reached a maximum level at 2 h and lasted up to 4 h thereafter (Fig. 6a). The delayed peaks of NF-κB due to phagocytosis were consistent with enhanced TNFα expression after phagocytosis (Fig. 1f). In addition, phagocytosis of IgG-coated beads alone did not activate NF-κB (Fig. 6b). These results indicate that phagocytosis enhances LPS-induced NF-κB activation in U937 cells.
To examine the effect of phagocytosis on the composition of the NF-κB complex, binding of the NF-κB complex was analyzed with antibodies against NF-κB subunits. Anti-p65 antibody and anti-p50 antibody shifted NF-κB bands activated by LPS but this did not occur with any other antibodies (Fig. 6c). Phagocytosis increased the shifted bands by anti-p65 as well as anti-p50 antibodies, and it did not induce any band shifts when probed with other antibodies including anti-RelB, anti-c-Rel, and anti-p52 antibodies, suggesting that phagocytosis did not change the composition of the NF-κB complex (Fig. 6c). Binding of the NF-κB complex simultaneously with both anti-p65 and anti-p50 antibodies further shifted the NF-κB band, confirming the presence of the p65/p50 heterodimer in the NF-κB complex (Fig. 6d). These results, taken together, indicated that phagocytosis increased LPS-induced activation of p65/p50. Furthermore, we examined the effect of phagocytosis on degradation of IκBα. IκB-α in cytoplasm decreased 15 min after LPS stimulation and almost disappeared 30 min after stimulation (Fig. 6e). IκB-α levels at these time points were comparable between cells that phagocytized beads and those that did not, indicating that phagocytosis did not promote the degradation of IκB-α (Fig. 6e). TLR2/6-mediated pro-inflammatory responses are also enhanced by phagocytosis. In order to determine whether the effects of phagocytosis are limited to stimulation with LPS, we sought to determine the effects of phagocytosis on macrophage-like cells stimulated with the synthetic diacyl lipopeptide, Palmitoyl-2-cysteine-serine-lysine-4 (Pam2CSK4), which is a ligand for the TLR2/6 heterodimeric receptor 22 . U937 cells stimulated with Pam2CSK4 were treated with IgG-coated beads, and expression of both TNFα and IL-6 was measured. Enhancement of cytokine expression by phagocytosis was observed; a 9.8-fold increase in TNFα expression and a 3.2-fold increase in IL-6 expression were observed from cells stimulated with 1000 ng/ml Pam2CSK4 (Fig. 7a,b). In addition, phagocytosis of IgG-coated beads increased NF-κB activation in U937 cells stimulated with Pam2CSK4, and the increase was evident 2 and 4 h after stimulation (Fig. 7c). Taken together, these results indicate that phagocytosis enhances Pam2CSK4-induced pro-inflammatory responses mediated by TLR2/6 signaling.

Discussion
In this study, we observed that phagocytosis of germ-free microparticles enhances the pro-inflammatory responses of human macrophage-like cells stimulated with bacterial components including LPS and lipopeptide. Although bacterial phagocytosis is reportedly important for macrophage response, we have demonstrated that phagocytosis itself influences the responsiveness of the phagocyte by using polystyrene latex beads. Various intracellular phenomena caused by phagocytosis have been proposed to exert positive effects on TLR signaling. Moreover, the uptake of bacterial components, including LPS and lipopeptide, has been thought to occur in association with phagocytosis, which leads to the enhancement of macrophage pro-inflammatory responses. In our experiments, bead phagocytosis caused the expression of cytokines to increase beyond the maximum level of cytokine expression stimulated by LPS or lipopeptide alone. This means that phagocytosis has an effect beyond ligand enrichment. Phagocytosis of particles is known to induce polarization of macrophages, such as the M1 and M2 phenotypes 23 . However, the enhanced effects of phagocytosis in our study do not seem to be a result of macrophage polarization; THP-1 cells polarize into the M1 or M2 phenotypes after treatment for 24-48 h with various stimuli 24 , but we treated cells simultaneously with beads and TLR ligands. Therefore, we believe that the phagocytosis signal acts either synergistically or additively on the ligand signal.  . (a,b) Activation of NF-κB in U937 cells was examined by EMSA. Arrowheads indicate the mobility-shifted bands. Full-length gels are presented in Supplementary Fig. 1. Cells stimulated with 0.1 ng/ml LPS were cultured simultaneously either with ( +) or without (−) beads pre-treated with 1000 µg/ml human IgG at a ratio of 1000 beads/cell for the indicated duration (a). Cells stimulated with ( +) or without (−) 1 ng/ml LPS were cultured simultaneously either with ( +) or without (−) beads pre-treated with 1000 µg/ml human IgG at a ratio of 1000 beads/cell for 2 h (b). (c,d) Binding of the NF-κB complex with indicated antibodies was examined by supershift assay. Arrowheads indicate the mobility-shifted bands and arrows indicate supershifted bands. Full-length gels are presented in Supplementary Fig. 1. Cells stimulated with 0.1 ng/ml LPS were cultured simultaneously either with (Beads +) or without (Beads −) beads pre-treated with 1000 µg/ml human IgG at a ratio of 1000 beads/ cell for 2 h (c). Cells were stimulated with 0.1 ng/ml LPS for 2 h (d). (e) Degradation of IκB-α was examined by western blot (upper panel). Proteins were analyzed by Coomassie brilliant blue staining (lower panel). An arrowhead indicates the IκB-α bands. A full-length blot is presented in Supplementary Fig. 1. Cells stimulated with 50 ng/ml LPS were cultured simultaneously either with (+) or without (−) beads pre-treated with 1000 µg/ ml human IgG at a ratio of 2,000 beads/cell for the indicated duration. www.nature.com/scientificreports/ There are some precedents that indicate crosstalk between TLR and the Fcγ receptor. Immune cells seeded on IgG-coated plate responded much more to LPS and/or lipopeptide than those plated on non-coated plate 25, 26 . In the present study, crosstalk between TLR and the Fcγ receptor does not explain the enhancement of TLR signaling by phagocytosis since phagocytosis of PLL-coated beads is not mediated by the Fcγ receptor. The cellular receptor for phagocytosis of PLL-coated particles has not been identified, but contact with polystyrene latex beads to cell membrane activates intracellular signaling that induces actin reorganization for phagocytosis of the beads 27 . Although intracellular signaling for phagocytosis differs depending on the mechanism of particle recognition 10,27,28 , actin reorganization may be induced in cells seeded on IgG-coated plates as well as in cells that phagocytize beads 29 . Therefore, actin reorganization may be involved in enhancing TLR signaling by phagocytosis. In addition, particle size is known to affect the processes of phagocytosis such as the uptake of particles and their intracellular trafficking 30 . Use of particles with various mean diameters may be valuable for the machinery for which phagocytosis influences LPS signaling.
In this study, phagocytosis enhanced both TLR4 and TLR2/6 signaling. TLR4 activates both the myeloid differentiation primary response 88 (MyD88) pathway and the TIR domain containing adaptor inducing interferon  . (a,b) U937 cells stimulated with the indicated concentrations of Pam2CSK4 were cultured simultaneously either with (gray bar) or without (white bar) beads pre-treated with 1000 µg/ml human IgG at a ratio of 500 beads/cell. After culturing, TNFα (a) and IL-6 (b) expression levels were determined. The data shown are the mean values with standard deviations derived from triplicate measurements. A dagger ( †) indicates values below the detection limit. Data were analyzed by the t-test. *P < 0.05. (c) U937 cells stimulated with 1 ng/ml Pam2CSK4 were cultured simultaneously either with (+) or without (−) beads pre-treated with 1000 µg/ml human IgG at a ratio of 1000 beads/cell for the indicated duration, and NF-κB activation was examined by EMSA. Arrowheads indicate shifted bands. A full-length gel is presented in Supplementary Fig. 1 www.nature.com/scientificreports/ β (TRIF) pathway whereas TLR2/6 only activates the MyD88 pathway [31][32][33][34] . Since phagocytosis enhances both TLR4 signaling and TLR2/6 signaling, phagocytosis has been proposed to enhance the common signaling pathway between TLR4 signaling and TLR2/6 signaling including the MyD88 pathway. The MyD88 pathway is the adaptor for downstream signaling by almost all TLRs [35][36][37] . Phagocytosis enhanced LPS-induced nuclear translocation of p65/p50 but minimally promoted IκB-α degradation. Although further analysis is required for understanding the mechanism that enhances p65/p50 activation by phagocytosis, phagocytosis seems to influence MyD88 signaling. If the phagocytosis signal acts on the MyD88 pathway, phagocytosis may have positive effects on signaling induced by almost all TLRs. However, the effects of phagocytosis on LPS-induced pro-inflammatory cytokines differed among cytokines. Phagocytosis induced robust expression of TNFα and IL-1β, but the effect on IL-6 expression was not robust. This difference in expression level may be related to differences in transcriptional regulation among cytokine genes. Transcription of the IL-6 gene with low CpG density in its promoter requires nucleosome remodeling to make it more accessible to transcription factors including NF-κB because of the otherwise condensed chromatin structure at the promoter region [38][39][40][41][42] . On the other hand, the TNFα gene, given the high CpG density in its promoter, may not require nucleosome remodeling [38][39][40][41][42] . Other cytokine genes that require nucleosome remodeling for their transcription, such as IL12B, may also show similar enhanced expression patterns compared with IL6 [40][41][42][43] . TLR signaling plays an important role for initiating the adaptive immune response 44 . Efficient TLR signaling is important for covering the immunogenicity of non-living vaccine antigens, which are often poorly immunogenic 45 . Our findings suggest that phagocytosis increases the adjuvant activity of TLR ligands. Previously, the biological activity of endotoxin has been reported to vary depending on the preparation method that affects particle size 46 . The particle size may have affected phagocytosis. To date, the phagocytosis of particulate adjuvants, including aluminum salts, are known to activate the nucleotide-binding domain and leucine-richrepeat-containing gene family pyrin-domain containing 3 (NLRP3) inflammasome, which processes pro-IL-1β and converts it into the mature form 47,48 . Combinatorial use of particulate adjuvants and TLR ligands may be useful for inducing potent immunogenicity 49,50 . For instance, aluminum salts have been used in combination with monophosphoryl lipid A 51-53 , a detoxified derivative of LPS [54][55][56] . Our findings suggest that particulate adjuvants can enhance TLR signaling in addition to the NLRP3 inflammasome, and demonstrate the effects of phagocytosis itself on TLR signaling. Combinatorial use of microparticle phagocytosis should be useful for many medical applications, such as regulating inflammation, potentiating the immune system, and vaccination. In this study we analyzed effects of phagocytosis on U937 and THP-1 cell lines, both of which were human monocytic cell lines. Both cell lines differentiated into macrophage-like cells exhibited increased responsiveness to LPS due to phagocytosis of microparticles. Therefore, we speculated that some types of macrophages should increase their responsiveness to LPS by phagocytosis under certain physiological conditions; however, the types of macrophages that show similar responses have not yet been determined. Further analysis to address this issue is important not only for understanding the immune system but also for the clinical application of this study.

Methods
Reagents. RPMI1640  Cell culture. U937 and THP-1 cells were cultured in RPMI1640 medium supplemented with 10% (v/v) heat-inactivated fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin under a humidified atmosphere with 5% CO 2 at 37 °C. To differentiate U937 and THP-1 cells into active macrophage-like cells, the cells were resuspended in culture medium containing 0.1 μM 1α, 25-dihydroxyvitamin D 3 to a density of 2.5 × 10 5 cells/ml. Resuspended cells were incubated at 37 °C for 3 d 57,58 . After differentiation, medium containing 1α, 25-dihydroxyvitamin D 3 was replaced with fresh culture medium, and the cells were used for analysis.
Treatment of polystyrene latex beads with human IgG or PLL. Beads were coated with IgG at the time of use. Polystyrene latex beads dyed blue with a mean diameter of 0.8 μm were suspended in water at 0.83% (w/w), and 750 μl of the suspension was centrifuged at 15,000 g for 10 min at 4 °C; the supernatant was subsequently removed. Precipitated beads were resuspended in 1 ml water before they were collected again by centrifugation. The procedure for washing with water was repeated three times. Number of the beads was calculated by equation provided by manufacture, and 3.57 × 10 9 beads were incubated in 400 µl RPMI1640 medium containing human IgG at 37 °C for 1 h. The supernatant was subsequently removed after centrifugation at 15,000 g for 5 min at 4 °C; the precipitated beads were suspended in 1 ml phosphate buffered saline (PBS) and www.nature.com/scientificreports/ the beads were collected by centrifugation at 15,000 g for 5 min at 4 °C. This procedure for washing with PBS was performed three times. Beads were resuspended in PBS at 8.93 × 10 10 beads/ml. PLL (300 µg/ml) was used to coat carboxylated polystyrene latex beads dyed blue with a mean diameter of 1.0 μm by covalent coupling using the carbodiimide method according to the manufacturer's protocol (Polysciences, Inc.). Prepared PLL-coated beads were stored at 4 °C until use.

Pre-treatment of cells with IgG and cytochalasin D for measuring phagocytosis. Human IgG
dissolved in 0.85% (w/v) NaCl solution was added to cells at final concentrations of 10 or 100 μg/ml 4 h prior to the addition of beads. Cytochalasin D dissolved in dimethylsulfoxide (DMSO) was added to cells to final concentrations of either 1 or 10 μM 1 h prior to addition of beads. The vehicle for each inhibitor was added to experimental controls. Cells were incubated at 37 ˚C until beads were added. Subsequent treatment of cells with beads was performed in the continuous presence of each inhibitor.
Evaluation of the number of phagocytized beads. The number of phagocytized beads was evaluated based on previously published methods, with slight modifications 57,59 . All subsequent manipulations were carried out at 4 °C or on ice unless otherwise noted. The cell suspension (1 ml) was layered over 1 ml of 100% NCS in a microcentrifuge tube. The layered cell suspension was centrifuged at 100 g for 10 min to separate cells from free beads, and the supernatant was discarded. The precipitated cells were resuspended in 1 ml PBS, and the cell suspension was layered over 1 ml of 100% NCS again. After centrifugation, the supernatant was removed. The procedure for separating cells from free beads using NCS was performed four times in total. Next, the separated cells were resuspended in 1 ml PBS and washed twice by centrifuging at 800 g for 10 min, and the washed cells were lysed with 500 µl 0.1% SDS solution at room temperature to release the beads from cells.  Electrophoretic mobility shift assay (EMSA) and supershift assay. EMSA was performed based on the methods published previously with slight modifications 62 . Nuclear extract (20 μg of protein amount) was incubated with 20 µl 0.01 μM NF-κB-specific oligonucleotide as described above in 5 mM Tris-HCl (pH 8.0) containing 50 mM NaCl, 0.05% Nonidet P-40, 1 mM MgCl 2 , 0.5 mM DTT, 5% (v/v) glycerol, 0.25 mM EDTA, 40 ng/ml BSA, and 0.2 mg/ml poly (deoxyinosinic-deoxycitidylic) acid sodium salt for 30 min at room temperature. Subsequently, the reaction mixture (15 µl) was loaded on 6% native polyacrylamide gels in 1 × Tris-borate-EDTA buffer (Nacalai Tesque, Kyoto, Japan). Electrophoresis was performed at 150 V for 60 min at 4 °C. Bands of 6-FAM-labeled DNA on the gels were detected with FLA-5100 (FUJIFILM Corporation, Tokyo, Japan) using an excitation wavelength of 473 nm and emission wavelength of 510 nm. Supershift assay was performed using antibodies against NF-κB subunits: p65, RelB, c-Rel, p50, and p52. Nuclear extract (20 μg of protein amount) was incubated with 5 μg of each antibody for 30 min at room temperature in a total volume of 10 µl. Vehicle control was incubated with PBS containing 0.1% (w/v) NaN 3 . EMSA was performed as described above except that the electrophoresis was performed for 70 min. www.nature.com/scientificreports/ Western blot. Proteins (50 µg) of cytoplasmic fractions were separated by sodium dodecyl sulfate-polyacrylamide electrophoresis and electroblotted on to a nitrocellulose membrane. After blocking with 5% (w/v) skimmed milk in tris-buffered saline (TBS) containing 0.1% (v/v) Tween 20 at room temperature for 1 h, the membrane was incubated with 1/1000 dilution of anti-IκB-α antibody in 5% (w/v) BSA in TBS-containing 0.1% (v/v) Tween 20 at 4 °C overnight. Horseradish peroxidase-conjugated antibody was used to develop the membrane and visualization of bands was performed using SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific, Inc.) with Amersharm Imager 600 (GE Healthcare Life Sciences).
Statistical analysis. Statistical significance was determined by Student's t-test using Statcel 3 (OMS publishing, Tokyo, Japan), where P < 0.05 was considered significant. If F-test did not assume the homogeneity of variance, Welch's t-test was used instead of Student's t-test.