Inorganic polyphosphate potentiates lipopolysaccharide-induced macrophage inflammatory response

Inorganic polyphosphate (polyP) is a linear polymer of orthophosphate units that are linked by phosphoanhydride bonds and is involved in various pathophysiological processes. However, the role of polyP in immune cell dysfunction is not well-understood. In this study, using several biochemical and cell biology approaches, including cytokine assays, immunofluorescence microscopy, receptor-binding assays with quartz crystal microbalance, and dynamic light scanning, we investigated the effect of polyP on in vitro lipopolysaccharide (LPS)-induced macrophage inflammatory response. PolyP up-regulated LPS-induced production of the inflammatory cytokines, such as tumor necrosis factor α, interleukin-1β, and interleukin-6, in macrophages, and the effect was polyP dose– and chain length–dependent. However, orthophosphate did not exhibit this effect. PolyP enhanced the LPS-induced intracellular macrophage inflammatory signals. Affinity analysis revealed that polyP interacts with LPS, inducing formation of small micelles, and the polyP-LPS complex enhanced the binding affinity of LPS to Toll-like receptor 4 (TLR4) on macrophages. These results suggest that inorganic polyP plays a critical role in promoting inflammatory response by enhancing the interaction between LPS and TLR4 in macrophages.

Inorganic polyphosphate (polyP) 2 is a linear polymer of orthophosphate units that are linked by phosphoanhydride bonds, similar to ATP (1). PolyP is ubiquitously found in all living cells, from microorganisms to mammals, and its chain length varies from less than 10 to hundreds of phosphate units (2). In humans, polyP, containing 60 -100 phosphate units, is stored in platelet granules and can be released into circulation upon its activation or other physiological stimuli (3). Recent studies have revealed that polyP exhibits various physiological effects, such as activating blood procoagulant cascades (4), eliciting pro-inflammatory response in endothelial cells (5,6), and promoting amyloid fibril formation (7,8), and it is possible that the reaction of polyP with inflammatory immune cells is associated with the pathology of infection (6,9,10).
Lipopolysaccharide (LPS) is a major component of the outer membrane of Gram-negative bacteria and is one of the primary endotoxins released by these organisms (11). LPS induces an inflammatory response in immunocompetent cells in human and mouse models (12,13), and it is associated with the incidence of both acute and chronic inflammatory diseases. Tolllike receptor 4 (TLR4) is an LPS receptor expressed on immune cells such as monocytes, macrophages, and granulocytes (12), and activation of TLR4-associated signaling, including mitogen-activated protein kinase (MAPK) and nuclear factor (NF)-B pathways, induces synthesis of inflammatory cytokines, such as tumor necrosis factor ␣ (TNF␣), interleukin (IL)-1␤, and IL-6 (13). However, the relation of LPS with polyP in the inflammatory response in immune cells is largely unknown. In this study, we investigated the effects of polyP on LPS-induced inflammatory response in macrophages, focusing on the reaction between polyP and LPS in vitro.

PolyP amplified LPS-induced inflammatory response in macrophages
To investigate the effect of polyP on LPS-induced macrophage inflammatory response, THP-1-derived macrophages were reacted with LPS isolated from Escherichia coli with or without polyP-65 (chain length 60 -70-mer). LPS induced the expression of inflammatory cytokine genes TNF␣, IL-1␤, and IL-6 in macrophages, as reported previously (14). The addition of polyP-65 markedly enhanced cytokine expression (TNF␣: polyP and LPS (7.16 Ϯ 1.98) versus LPS alone (1.14 Ϯ 0.55); IL-1␤: polyP and LPS (5.70 Ϯ 1.49) versus LPS alone (1.38 Ϯ 0.22); IL-6: polyP and LPS (16.11 Ϯ 8.63) versus LPS alone (0.92 Ϯ 0.60), p Ͻ 0.05), whereas polyP-65 itself did not induce macrophage inflammatory response in the absence of LPS (Fig.  1, A-C). Consistent with gene expression, polyP-65 greatly amplified the production of cytokines from macrophages in a dose-dependent manner (Fig. 1, D-F). The LPS-induced The authors declare that they have no conflicts of interest with the contents of this article. This article contains Figs. S1-S6. 1 To whom correspondence should be addressed. cro ARTICLE macrophage inflammatory cytokine production was also enhanced in a chain length-dependent manner within a range of 1-850 mer (Fig. 1, G-I). PolyP did not involve in cell viability in the reaction setting (Fig. S1). Cytokine production was also increased in human peripheral blood mononuclear cells and J774.1 murine macrophages (Fig. S2). These results indicate that polyP amplified LPS-induced macrophage inflammatory response in a dose-and chain length-dependent manner.

PolyP amplified macrophage inflammatory response via downstream signaling of TLR4
To examine the mechanism of polyP-mediated amplification of LPS-induced macrophage inflammatory response, intracellular signaling molecules related to inflammatory response were assessed. PolyP did not change the mRNA expression of TLR4 and the components of the complex, such as CD14 and myeloid differentiation protein-2 (MD-2) in THP-1 cells reacted with LPS (Fig. S3). LPS induced phosphorylation of JNK, NF-B, and IB␣, but not p38, Erk1/2-MAPK. The phosphorylation was enhanced with the addition of polyP-65 (Fig. 2, A-E). In contrast, polyP-65 alone did not affect phosphorylation of these proteins. Furthermore, polyP-65 amplified LPSinduced NLRP3 protein expression in macrophages (Fig. 2F).
These results suggest that polyP enhances LPS-induced MAPK, NF-B, and NLRP3 activation in macrophages.
To elucidate the mechanism of polyP-mediated enhancement of LPS-induced activation of downstream signaling of TRL4 in macrophages, we performed experiments using a receptor and a signal blocker associated with LPS. Both TAK242, an antagonist of TLR4, and BAY11-7082, an inhibitor of NF-B, significantly inhibited the synergistic effect of polyP on LPS-induced cytokine synthesis (Fig. 3, A-F).

PolyP promoted binding of LPS with TLR4 on macrophages
To investigate the molecular interactions among polyP, LPS, and TLR4 on macrophages, we used immunofluorescent microscopy, flow cytometry, and quartz crystal microbalance (QCM) analysis. When macrophages were incubated with biotinylated LPS, several dot signals emerged on the surface of cells, and polyP-65 enhanced biotinylated LPS staining on the surface of macrophages (fluorescence intensity: polyP with LPS (213.8 Ϯ 55.7 A.U.) versus LPS alone (11.7 Ϯ 6.3 A.U.), p Ͻ 0.01) (Fig. 4, A and B). Flow cytometry analysis showed that treatment with polyP significantly increased the abundance of LPS-FITC-positive macrophages compared with those treated with LPS alone (mean intensity: polyP with LPS (45.62 Ϯ 1.07 A.U.) versus LPS alone (26.11 Ϯ 0.36 A.U.), p Ͻ 0.01) (Fig. 4, C and D). Furthermore, to assess the effect of polyP on the interaction between LPS and TLR4, we performed QCM analysis (Fig. 4, E-J). A decrease in resonance frequency was observed when LPS was injected onto a sensor chip containing immobilized TLR4 (Fig. 4G). In this system, injection of polyP-65 plus LPS led to a higher reduction in the frequency than was observed after injection of LPS alone (Fig. 4, H-J), whereas injection of polyP-65 or buffer alone did not change the frequency (Fig. 4, E and F). To calculate the kinetic parameter, repetitive injection of LPS and polyP-65 was performed. The amount of change in frequency for each injection was greater in LPS with polyP-65 than in polyP-65 alone. (Fig. S4). The K D values of LPS to TLR4 were 2.07 M (LPS) and 1.51 M (LPS ϩ polyP-65), respectively. We also examined the effect of polyP on the interaction between LPS and other TLR4 complexes, such as CD14 and MD-2, while there was no interaction with LPS (Fig. S5). These results indicate that polyP enhances the interaction between LPS and TLR4 on macrophages.

Interaction between polyP and LPS promoted the formation of smaller LPS micelles
To obtain additional information on the interaction between polyP and LPS, we measured the hydrodynamic diameter (D h ) of LPS micelles in the presence of varying concentrations of polyP-65 using dynamic light scattering (DLS) (Fig. 5A). A pre-vious study indicated that LPS forms distinct types of micelles, depending on the concentrations (15); specifically, premicelle oligomers and large aggregates were formed below and above, respectively, the apparent critical micelle concentration of 14 g/ml LPS. The DLS measurement of LPS micelles in the absence of polyP-65 detected two primary components: com- To examine the affinity between polyP and LPS, we performed isothermal titration calorimetry (ITC) experiments. The titration of polyP-65 with LPS showed endothermic heat, with a stronger interaction at 25°C than at 37°C (Fig. 5D). As a control, endothermic heat was not observed with LPS alone (Fig. S6A) or polyP-65 alone (Fig. S6B). We then analyzed the saturating titration curve at 25°C (Fig. 5E). Parameters could not be assessed at 37°C owing to poor curve fitting. The

Polyphosphate and macrophages
apparent stoichiometry of binding was as follows: 1 mol of LPS interacted with 0.036 -0.072 mol of polyP-65 (14 -28 mol of LPS interacted with 1 mol of polyP-65) when the molecular weight of LPS was assumed to be in the range of 3,000 -6,000, with a K d of 0.26 Ϯ 0.05 M. The ⌬H, ϪT⌬S, and ⌬G values for binding were 199 Ϯ 111, Ϫ208, and Ϫ8.81 kcal/mol, respectively. Thermodynamic parameters indicated that the driving force for favorable ⌬G arises from the positive entropy change (⌬S). These results suggest that polyP-65 interacts with LPS micelles consisting of 14 -28 molecules, leading to a net dehydration (i.e. removal of water molecules) around polyP-65 or LPS micelles, and converts the large micelles into smaller micelles.

Discussion
In this study, we report that inorganic polyP amplified LPSinduced macrophage inflammatory response in vitro by enhancing the binding affinity of LPS to TLR4 and promoting LPS micelle formation.
Several studies have shown the role of polyP in inflammation and blood clotting cascades. For example, polyP amplifies the inflammatory response of nuclear proteins in human umbilical vein endothelial cells by enhancing multiple ligand-receptor signaling pathways (6,10). In our study, polyP significantly enhanced LPS-induced cytokine production in macrophages in a dose-and chain length-dependent manner (Fig. 1). In this experimental system, monophosphate did not promote macrophage inflammation, suggesting that polymerization of phosphate could be crucial for the reaction. The effect of polyP on LPS-induced macrophage inflammation was more prominent when the chain length was greater than 65 monomers (Fig. 1,  G-I). Middle-length chain polyP (60 -100-mer) is known to be stored in platelets and released into circulation during systemic

Polyphosphate and macrophages
inflammation (4,16). Long-chain polyP (greater than several hundred monomers) is synthesized enzymatically and stored in bacteria (17) and is released into circulation during infection (6, 10, 17). Thus, it is possible that circulating middle-and/or longchain polyP was produced during sepsis and enhanced the reaction between LPS and immune cells.
To understand the detailed mechanism of the effect of polyP on LPS-induced macrophage inflammatory response, we examined the reaction of polyP with LPS, which then reacted with TLR4 on the surface of immune cells. DLS measurements indicated that polyP induced the conversion of larger LPS micelles to smaller micelles in a polyP concentration-dependent manner (Fig. 5B). ITC measurements indicated that 14 -28 mol of LPS interacted with 1 mol of polyP-65 by an entropy-driven reaction. LPS consists of three components, lipid A, O-specific chain, and core oligosaccharide, (11,15), and has been reported to form micelles (15). Our QCM analysis and immunofluorescence studies (Figs. 4A and 5) revealed that the negatively charged polyP enhanced the interaction between TLR4 and LPS that accelerated micelle formation. Previous reports have shown that polyP binds to nuclear cytokines and stabilizes ligand-receptor affinity (6, 10); however, our experimental system showed that polyP enhanced the interaction between LPS and macrophages, but polyP itself did not induce macrophage inflammatory response (Figs. 1-3).
Based on our findings, we proposed a novel hypothesis for the role of polyP in LPS-induced macrophage inflammatory response (Fig. 6). PolyP is stored in serum, plasma, or platelets of individuals. Sepsis induces the release of polyP and LPS from bacteria or activates platelets to release their stored polyP. The released or stored polyP promotes LPS to form smaller micelles, which facilitate binding of LPS to TLR4 on immune cells with higher intensity than in the absence of polyP. Thus, the polyP-LPS complex amplifies the signaling from TLR4 to downstream

Polyphosphate and macrophages
molecules, such as MAPK and NF-B, which enhance inflammatory cytokine production from macrophages. Excess cytokine production driven by polyP may exacerbate systemic inflammatory disease. Our results are limited to in vitro experiments, and further research is needed to elucidate the role of polyP and its metabolism in clinical sepsis.
In conclusion, polyP amplified LPS-induced macrophage inflammatory response in vitro. PolyP promoted smaller micelle formation by LPS and accelerated the binding of LPS to TLR4 on macrophages. Our findings provide new perspectives and a novel therapeutic target for inflammatory diseases.

Detection of proteins in cells
THP-1 macrophages were incubated with or without LPS (1.0 ng/ml) and/or polyP-65 (50 M) for 6 h. Western blot analysis was performed to detect the phosphorylation of JNK, p38, Erk1/2-MAPK, NF-B p65, and IB␣ and the expression of NLRP3 in cells. Whole-cell lysates were lysed with radioimmune precipitation assay buffer (FUJIFILM Wako, Osaka, Japan), supplemented with protease inhibitor mixture (Roche, Basel, Switzerland) and phosphatase inhibitor mixture (Roche). The lysates were centrifuged at 13,000 ϫ g for 20 min at 4°C, and the protein concentration was evaluated using a BCA protein assay kit (Thermo Fisher Scientific) according to the manufacturer's protocol. The proteins were boiled in SDS sample buffer (TEFCO, Tokyo, Japan) and 10% 2-mercaptoethanol at 95°C for 5 min. The sample was resolved using 10% SDS-PAGE and transferred to a polyvinylidene fluoride membrane (Atto, Tokyo, Japan). Membranes were blocked with Western Blocking Reagent (Roche), which was diluted with 50 M Tris-buffered saline (TBS), and incubated with primary antibodies overnight at 4°C. After washing with TBS containing 0.05% Tween 20, the membranes were incubated with HRP-linked secondary antibodies for 1 h at 20°C. After incubation with secondary antibodies, signals were visualized as chemiluminescence by ECL Western blotting substrate (Thermo Fisher Scientific) and detected using WSE-6100 LuminoGraph (Atto). Chemiluminescence intensity was calculated using ImageJ software (National Institutes of Health, Bethesda, MD). All antibodies were diluted 1:1,000.

Cell viability assay
To assess the viability of THP-1 macrophages reacted with LPS and polyP-65, we performed a cell viability assay using Cell Counting Kit-8 (CCK-8) (DOJINDO, Kumamoto, Japan) according to the manufacturer's protocol (18).

Reaction with NF-B inhibitor and TLR4 receptor inhibitor
THP-1 macrophages were incubated with a 1.0 M concentration of either the NF-B inhibitor BAY11-7082 (Sigma-Aldrich) or the TLR4 inhibitor TAK-242 (Sigma-Aldrich) for

Polyphosphate and macrophages
1 h before incubation with LPS and polyP. After incubation, levels of cytokines in cell supernatant were measured using ELISA kits as described above.

Immunofluorescence microscopy
THP-1 cells were seeded in 8-well culture slides (Corning, Inc., Corning, NY) and differentiated to macrophages as described above. Differentiated cells were incubated with or without 1.0 g/ml biotinylated LPS (In vivo Gen, San Diego, CA) and/or polyP-65 (50 M) for 3 h. Cells were then washed with PBS and fixed with 4% paraformaldehyde for 20 min. After washing with PBS, cells were incubated with FITC-labeled antibiotin rabbit polyclonal antibody (Abcam, Cambridge, UK) for 1 h at 20 -25°C (diluted 1:500). After washing with PBS, the cells were incubated with rhodamine phalloidin (Thermo Fisher Scientific) to visualize F-actin. Stained cells were mounted with HardSet Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA). Images were acquired using a FV1200confocalmicroscope(Olympus,Tokyo,Japan).Thefluorescence intensity was quantified using ImageJ software.

Flow cytometry
THP-1 macrophages were incubated with biotinylated LPS with or without polyP-65 (50 M) for 3 h. Cells were then washed with PBS and incubated with FITC-labeled anti-biotin rabbit polyclonal antibody for 1 h at room temperature (diluted 1:500). After washing with PBS, cells were collected and washed with FACS buffer (3% FBS and 0.01% sodium azide in PBS). Flow cytometric analysis was performed with FACSCalibur using Cell Quest Pro software (BD Biosciences). Analysis was performed after 10,000 counting events. The mean fluorescence intensity value was used to estimate LPS levels in cells.

Assessing interaction between LPS and TLR4, MD2, and CD14 using QCM
Binding of LPS to TLR4 was assessed using a highly sensitive 27-MHz instrument (AFFINIX Q8, ULVAC, Kanagawa, Japan) as described previously (19). Briefly, QCM sensor chips were prepared using the immobilization kit for AFFINIX (ULVAC) according to the manufacturer's protocol. Recombinant human TLR4 (50 g/ml), MD2 (50 g/ml), and CD14 (50 g/ml) (R&D Systems) were then immobilized on the sensor chip. The sensor chips were soaked in distilled water in the incubation chamber. LPS was injected into the chamber in the presence or absence of polyP-65 for binding to TLR4. The resonance frequency of QCM at equilibrium was defined as the zero position. The stability and drift of the 27-MHz QCM frequency in solution were Ϯ3 Hz. Binding affinity was determined by the frequency change following injection of each sample.
Kinetic analysis between LPS and polyP-65 was performed using repetitive injection of LPS and LPS with polyP-65 as described previously (20). AQUA software (ULVAC) was used to calculate the kinetic parameters.

Dynamic light scanning
DLS measurements were performed with 0.01 mg/ml LPS solutions containing varying concentrations of polyP-65, 10 mM NaP i (pH 7), and 150 mM NaCl at 37°C using Zetasizer V (Malvern Panalytical, Worcestershire, UK). Intensity distribution data were mathematically converted into volume distribution data, which were fitted by Gaussian curves to evaluate the fraction of distinct components.

Isothermal titration calorimetry
ITC measurements were performed using a VP-ITC instrument (MicroCal, Northampton, MA) at neutral pH (H 2 O) at 25°C and 37°C. Five microliters of 20 M polyP-65 in a syringe was titrated into 0.01 mg/ml LPS in the ITC cell. For control experiments, 5 l of 20 M polyP-65 was titrated into H 2 O, or 5 l of H 2 O was titrated into 0.01 mg/ml LPS. During ITC experiments, 20 successive injections were performed 180 s apart, and the cell was continuously stirred at 633 rpm. ITC data were analyzed using Origin software. The thermodynamic parameters were obtained from the integrated heat using a single set of binding sites model.

Statistical analysis
Data were expressed as mean Ϯ S.D. Statistical analysis was performed by Student's t test when only two value sets were compared. One-way analysis of variance (ANOVA) followed by Bonferroni's multiple-comparison test was used to compare three or more groups. The number of biological samples (n) for each group and the utilized statistical tests are indicated in the corresponding figure legends. Differences in means were considered statistically significant if p Ͻ 0.05. No exclusion criteria were incorporated in the design of the experiments for this study.