Multi-walled carbon nanotubes act as a chemokine and recruit macrophages by activating the PLC/IP3/CRAC channel signaling pathway

The impact of nanomaterials on immune cells is gaining attention but is not well documented. Here, we report a novel stimulating effect of carboxylated multi-walled carbon nanotubes (c-MWCNTs) on the migration of macrophages and uncover the underlying mechanisms, especially the upstream signaling, using a series of techniques including transwell migration assay, patch clamp, ELISA and confocal microscopy. c-MWCNTs dramatically stimulated the migration of RAW264.7 macrophages when endocytosed, and this effect was abolished by inhibiting phospholipase C (PLC) with U-73122, antagonizing the IP3 receptor with 2-APB, and blocking calcium release-activated calcium (CRAC) channels with SK&F96365. c-MWCNTs directly activated PLC and increased the IP3 level and [Ca2+]i level in RAW264.7 cells, promoted the translocation of the ER-resident stromal interaction molecule 1 (STIM1) towards the membranous calcium release-activated calcium channel modulator 1 (Orai1), and increased CRAC current densities in both RAW264.7 cells and HEK293 cells stably expressing the CRAC channel subunits Orai1 and STIM1. c-MWCNTs also induced dramatic spatial polarization of KCa3.1 channels in the RAW264.7 cells. We conclude that c-MWCNT is an activator of PLC and strongly recruits macrophages via the PLC/IP3/CRAC channel signaling cascade. These novel findings may provide a fundamental basis for the impact of MWCNTs on the immune system.

To determine whether RAW264.7 macrophages can phagocytize c-MWCNTs, we performed a transmission electron microscopy (TEM) study on the macrophages collected from the lower face of the PET membrane of the transwell insert (representing the migrated RAW264.7 cells). The TEM images ( Figure S1) show that c-MWCNTs were present inside the RAW264.7 cells exposed to c-MWCNTs for 24 h, and most of them were likely wrapped by a vesicle-like structure, but a few of them directly pierced the cytoplasm. This result clearly shows that RAW264.7 macrophages can phagocytize c-MWCNTs in our experimental system. c-MWCNTs activate PLC and increase the IP3 level in RAW264.7 macrophages. As the macrophage-recruiting effect of c-MWCNTs can be abolished by PLC inhibitor U-73122 as shown in Fig. 1E, we speculated that PLC may be the initial molecule of the signaling chain by which c-MWCNTs stimulate macrophage migration. Consistent with this speculation, the ELISA assay showed that c-MWCNT exposure for 6 h significantly increased the PLC activities in the supernatant of RAW264.7 cell lysates in which c-MWCNTs were removed by centrifugation at 12, 000 rpm, and this effect was abolished by PLC inhibitor U-73122 in the absence of c-MWCNTs (Fig. 1J). These results strongly suggest that c-MWCNTs can directly activate PLC. Accordingly, activation of PLC resulted in an increase of the IP3 level and this effect was also inhibited by U-73122 (Fig. 1K). These results suggest that c-MWCNTs may trigger calcium mobilization by activating the PLC/IP3/IP3R signaling.
In addition, we investigated the effect of c-MWCNTs on the mRNA level with real-time quantitative PCR and found that c-MWCNT exposure for 6 h significantly increased the mRNA expression of PLC ( Figure S2). c-MWCNTs activate CRAC channels and induce calcium mobilization in RAW264.7 macrophages. Based on the calcium mobilization hypothesis of c-MWCNTs mentioned above and the finding that blockage of CRAC channels abolished the macrophage-recruiting effect of c-MWCNTs as shown in Fig. 1G, we examined the potential effects of c-MWCNTs on the CRAC channel currents (I CRAC ) using the whole-cell configuration of a patch clamp and measured the intracellular free calcium level ([Ca 2+ ] i ) using confocal microscopy in RAW264.7 macrophages.
As DMEM was used to disperse c-MWCNTs, we also observed whether the pipette solution containing 50 μl/ ml DMEM (the final concentration used for dispersing c-MWCNTs in the pipette solution) without c-MWCNTs would exert an effect on I CRAC . The result showed that 50 μl/ml DMEM did not significantly affect I CRAC compared with the pipette solution without DMEM ( Figure S3) . The results of confocal microscopy showed that c-MWCNTs significantly increased the [Ca 2+ ] i level in RAW264.7 cells (Fig. 3). As a positive control drug, TG also increased the [Ca 2+ ] i level in RAW264.7 cells (Fig. 3).
We also examined whether c-MWCNTs would exert an effect on the production of reactive oxygen species (ROS) in RAW264.7 macrophages. The result ( Figure S4) showed that c-MWCNTs (50 μg/ml) moderately increased the ROS level. As a positive control agent in inducing ROS, H 2 O 2 (0.5 mmol/L) strongly increased the ROS level in RAW264.7 macrophages. c-MWCNTs stimulate the translocation of STIM1 to the PM in HEK293 cells and RAW264.7 cells. Using confocal microscopy, we demonstrated the translocation of STIM1 upon c-MWCNT exposure in HEK293 cells stably expressing EYFP-STIM1 and CFP-Orai1 and in RAW264.7 cells. In HEK293 cells, Orai1 was consistently in the PM (Fig. 4), and STIM1 was mainly distributed in the cytoplasm before exposure to c-MWCNTs (STIM1 is theoretically ER-resident at resting condition). There was a very low level of STIM1 in the PM in resting HEK293 cells, but it was diffusely distributed in the cytoplasm (Fig. 4, upper row). Upon c-MWCNT (50 μg/ml) stimulation for 24 h, STIM1 were obviously translocated to the PM area, as the STIM1 fluorescence signal in the PM area was significantly increased (Fig. 4, middle row). As a known ER-depleting agent, TG (1 μmol/L) exposure for 30 min strongly redistributed the STIM1 to the PM area (Fig. 4, lower row).
Similar to the HEK293 cells, in resting (control) RAW264.7 cells, STIM1 was diffusely distributed in the cytoplasm (Fig. 5, upper row). As the ER could not be clearly recognized under a confocal microscope, it is assumed that STIM1 was diffusely distributed in the ER membrane in this condition. Exposure of RAW264.7 cells to c-MWCNTs (50 µg/ml for 24 h) led to a dramatic translocation of STIM1 to the PM area, and STIM1 distribution in the PM presented a discrete punctate manner (Fig. 5, lower row). The co-localization images of STIM1 and nuclei (DAPI stained) also support the translocation of STIM1 to the PM (Fig. 5). The above results suggest that c-MWCNTs stimulate the translocation of diffusely ER-resident STIM1 to the ER-PM junctions in HEK293 cells and RAW264.7 cells. c-MWCNTs induce spatial polarization of the KCa3.1 channel in RAW264.7 cells. In the control RAW264.7 cells (before exposure to c-MWCNTs), most of the KCa3.1 channels were uniformly and diffusely distributed in the cytoplasm and some were likewise distributed in the PM (Fig. 6, upper row). However, after exposure to c-MWCNTs, the KCa3.1 channels were strikingly assembled to one pole of the RAW264.7 cells (Fig. 6, lower row). These results suggest that c-MWCNTs activate the KCa3.1 channels and induce spatial polarization of these channels, a phenomenon suggesting a role of the KCa3.1 channel in the process of polarity and steering behavior during macrophage migration.

Discussion
The present study focused on the impact and mechanism of c-MWCNTs on macrophage migration. Based on the results of transwell assays, we found that c-MWCNTs act as a chemokine and strongly stimulate the migration of RAW264.7 macrophages. We further outlined the underlying upstream signaling pathways (summarized in Fig. 7), i.e., after internalization, c-MWCNTs first activated PLC, thus leading to elevation of the IP3 level in the macrophages; IP3 induced Ca 2+ release from the ER by activating the Ca 2+ -releasing channel IP3R and thus led to ER depletion; the ER-resident STIM1 sensed the ER depletion and then translocated from the ER to the ER-PM junctions, where it reconstituted the complete CRAC channels with the membranous Orai1 and led to the  CRAC channel opening and Ca 2+ influx; Ca 2+ release and influx resulted in Ca 2+ mobilization and macrophage activation, then triggering the subsequent downstream signaling and driving the macrophages to undergo certain immunological activities such as migration and secretion. At the same time, the c-MWCNT-induced Ca 2+ mobilization also activated the KCa3.1 channel and accordingly led to K + efflux and a certain degree of membrane hyperpolarization, and this effect might restrict the macrophage activities to an appropriate level to avoid excessive immune responses. These findings might provide a fundamental basis for the impact of MWCNTs on macrophages and potentially even other immune cells and imply that the chemokine-like action of MWCNTs is important for further nanobiomedical practices utilizing carbon nanomaterials, both for curative and adverse effects evaluations. Now that c-MWCNTs can recruit macrophages, we wondered what the first targeting signaling molecule for c-MWCNTs would be to initiate the migration event. By observing the reaction time course of RAW264.7 macrophages to c-MWCNT stimulation, we realized that c-MWCNTs must enter the macrophages to act. For example, in the patch clamp study, extracellular application of c-MWCNTs could not increase I CRAC within the observed 20 min, while direct intracellular application of c-MWCNTs via the pipette solution soon increased I CRAC . We have observed similar phenomena in studying the effects of c-MWCNTs on the Kv4.2/4.3 channels in HEK293 cells and rat ventricular myocytes 10 . Extracellular application of c-MWCNTs for a longer time (for example >30 min) can increase I CRAC , suggesting that the time latency is the time needed for phagocytosis of c-MWCNTs. We also noticed in a previous study 9 that c-MWCNTs in the PC12 cells also needed a longer time to significantly reduce the current densities of three types of K + channels. These studies suggest that c-MWCNTs act at the intracellular side, and of course internalization of c-MWCNTs needs time.
The question remains: What is the first targeting signaling molecule of c-MWCNTs at the intracellular side of macrophages? We speculated and experimentally demonstrated that it is PLC, an enzyme that is the most upstream signaling molecule in controlling the Ca 2+ mobilization in non-muscle cells, especially for explicating the action mechanism of c-MWCNTs, as c-MWCNTs are unlikely to be equipped with specific membrane receptors that react to extracellularly acting agonists. We showed that PLC inhibitor U-73122 almost totally inhibited the c-MWCNT-induced macrophage migration, a result strongly supporting the above speculation. Some other investigators have provided clues to the effects of CNTs on PLC signaling. For example, Guidetti et al. 24 showed that CNTs could induce platelet activation in vitro through stimulation of PLC/Rap1b/integrin α(IIb)β3 signaling pathways; Matsumoto et al. 25 showed that CNTs stimulated neurite outgrowth of NGF-treated neurons through the PLC/PKC/ERK signaling pathway independent of the Ras/Raf/MEK cascade. Here, we demonstrated that c-MWCNTs induced macrophage migration by targeting PLC as the most upstream signaling molecule. Meng et al. 16 showed that MWCNTs stimulate the secretion of some chemokines such as macrophage inflammatory protein (MIP)-1α and MIP-2 that can recruit RAW264.7 macrophages. We thought that the chemokine secretion they observed was actually the downstream signaling in the event of MWCNT-induced migration of RAW264.7 cells. The present study mainly depicted the upstream signaling by which c-MWCNTs induced migration of RAW264.7 cells.
An additional important question emerges: Is c-MWCNT an activator of PLC or a nanozyme with PLC activity? In recent years, some nanomaterials, including carbon nanomaterials, have been found to exhibit unexpected enzyme-like activities (nanozymes) and have been considered the next-generation artificial enzymes 26,27 . Although the abovementioned studies 24,25 suggest that CNTs may activate PLC, the researchers did not measure the activities of PLC of cell lysates in their studies, and thus could not answer the above question. In the present study, we showed that U-73122, a PLC inhibitor, almost totally abolished the c-MWCNT-induced migration of RAW264.7 cells (see Fig. 1E). We further identified that incubation of RAW264.7 macrophages with c-MWCNTs for 6 h directly increased the PLC activity by up to fourfold (shown in Fig. 1J) in the supernatant of cell lysates in which c-MWCNTs had been removed by centrifugation at 12, 000 rpm, and the absence of c-MWCNTs in the supernatant was confirmed by light microscopy and TEM. These studies strongly suggest that c-MWCNT is an activator of PLC, rather than an artificial nanozyme possessing the PLC activity. Overall, we identified for the first time that c-MWCNTs induce macrophage migration by activating the PLC/IP3/CRAC channel signaling cascade in RAW264.7 macrophages, and PLC is the most upstream target signaling molecule for c-MWCNTs to induce macrophage migration, and importantly, c-MWCNT is an activator of PLC. We speculate that CNTs are possibly universal PLC activators, although it still needs to be investigated whether all types of CNTs are PLC activators. This PLC-activating character of c-MWCNTs may have broad use in many industrial fields, including enzymatic degumming of plant oils [28][29][30] , manufacturing of biodiesels, food, nutraceuticals, detergents, leather, paper, cosmetics and even agriculture 31 . In biomedical fields, PLC and/or its manipulators may have applications in bioremediation 31 , biosensoring 32 , and microenvironmental modeling of tumors through macrophage recruitment 33,34 . In addition, the potential for killing tumor cells using tumor-targeting MWCNTs to recruit certain type(s) of macrophages in the tumor sites merits investigation, although most recent studies on the role of macrophages in tumor development focus on the tumor-promoting effects of tumor-associated macrophages 35,36 .
Except for the finding of the PLC/IP3/CRAC channel signaling cascade as a "forward" pathway by which c-MWCNTs induce macrophage migration, we also identified the involvement of the KCa3.1 channel, which may potentially act as a "feedback" signal in the c-MWCNT-induced behavioral event of macrophages, as the function of the KCa3.1 channel is to introduce K + efflux when activated by Ca 2+ , and it thus provides a feed-forward effect on the initial influx of Ca 2+ and preserves the negative membrane potential required for sustained Ca 2+ influx 37 . In addition, KCa3.1 not only acts as a feedback signal for the behaviors of macrophages, but it may also facilitate the migration of macrophages cooperating with other channels such as Kv1.3 and CRAC channels. The dramatic cellular polarization of KCa3.1 in RAW264.7 macrophages that we observed in response to c-MWCNT exposure suggests the activation of the KCa3.1 channel and its certain roles in the migration process of macrophages, potentially including the formation of protruding and cell polarity. KCa3.1 has been found to promote the migration of other types of cells, including human lung mast cells 38 , human fibrocytes 39 and mouse neuroblasts in vivo 40 . The present study further identified the involvement of KCa3.1 in the c-MWCNT-induced migration of RAW264.7 macrophages. However, the exact role of KCa3.1 in c-MWCNT-induced macrophage migration needs further investigation.
Although we demonstrated the key upstream signaling by which c-MWCNTs triggered the macrophage migration, some other effects of c-MWCNTs on the macrophages, such as the enhancement of PLC gene transcription and ROS production that we showed, might be involved in the migration event or other activities of macrophages.
In summary, the present study demonstrated that c-MWCNTs act as a chemokine and strongly stimulate the migration of RAW264.7 macrophages by activating the PLC/IP3/CRAC signaling pathway. In particular, PLC serves as the most upstream targeting signaling molecule of c-MWCNTs, and c-MWCNT is an activator of PLC. The KCa3.1 channel is also involved in this event. These findings may shed light on the biomedical impact of CNTs at both the theoretical and practical level.

Preparation and characterization of MWCNTs. We used the carboxylated MWCNTs (c-MWCNTs)
in this study, as this type of CNT has a good hydrophilic property and can disperse well in the aqueous phase. The c-MWCNTs were developed from the pristine MWCNTs (p-MWCNTs). The p-MWCNTs were purchased from Chengdu Organic Chemicals Co. Ltd. (Chengdu, China) and the c-MWCNTs were prepared as previously described 41 . In brief, the c-MWCNTs were synthesized based on the p-MWCNTs through a combined oxidation procedure and probe sonication. The as-received MWCNTs were dried at 50 °C in a vacuum oven overnight and then were suspended in a 3:1 (volume) mixture of concentrated H 2 SO 4 /HNO 3 and sonicated with a power of 750 W for 80 sec. The resulting mixture was diluted with a large amount of distilled water, filtered and rinsed thoroughly until the pH was neutral before drying to constant weight at 50 °C in a vacuum oven. The prepared c-MWCNTs were subjected to scanning electron microscopy (SEM, Hitachi S-5200) to characterize their morphology. The length distribution for the c-MWCNTs was obtained by counting more than 300 nanotubes randomly taken in ten SEM images. X-ray photon spectroscopy (XPS, Japan JEOL Scientific JPS-9010TR) and Fourier transform infrared spectroscopy (FTIR, Necolet NEXUS 670) were applied to analyze their surface chemistry. The powder sample of c-MWCNTs was sterilized by autoclaving and then dispersed in Dulbecco's modified Eagle's medium (DMEM) to obtain a stock solution of 1 mg/ml by the aid of probe sonication (sonication time: 60 sec; working mode: working 3 s following 3 s stop; working power: 600 watt). For cell studies, c-MWCNTs were diluted in DMEM at a concentration of 50 μg/ml. The physicochemical characteristics of c-MWCNTs were previously described 42 . The average length of pristine MWCNTs was 50 μm. The length distribution for c-MWCNTs Scientific RepoRts | 7: 226 | DOI:10.1038/s41598-017-00386-3 ranged from 300 nm to 1.5 μm, and the average length was 926-945 nm depending on the manufacturing batch. The FTIR spectrum of c-MWCNTs showed that the characteristic absorption of the carboxylic group at 1720 cm −1 is substantially decreased in the c-MWCNT spectrum, while an absorption peak at 1630 cm −1 appeared.

Evaluation of potential endotoxin contamination in the MWCNTs.
To inactivate the potentially contaminated endotoxin in the c-MWCNT samples, the above prepared c-MWCNTs were routinely autoclaved (temperature >250 °C) and were again heated in an oven at 250 °C for 45 min prior to the experiment, with the suggested procedures of the International Pharmacopoeia (Sixth Edition, 2016) as a reference. The potential presence of endotoxin in the c-MWCNT solution was further evaluated using the limulus amebocyte lysate (LAL) gel clot assay kit (sensitivity 0.125 EU/ml) (Houshiji Company, Xiamen, China). To perform the assay, c-MWCNTs were added into the endotoxin-free water, vortexed and shaken for 15 min, followed by centrifugation at 12,000× g for 10 min. The supernatant was collected and added to the LAL assay solution (v/v = 1:1). The mixture was incubated at 37 °C for 1 h. Then, the tube was inverted to observe gel clot formation. Negative results were obtained with the c-MWCNT solution, demonstrating that the c-MWCNTs we used were endotoxin-free. The endotoxin of 0.25 EU/ml was set as a positive control which could induce gel clot formation. Because c-MWCNTs produced a black solution, we did not use a colorimetric test to examine endotoxin contamination in the c-MWCNT solution, out of concern that it may yield a false-positive result.
Cell culture and transfection. The RAW264.7 macrophage-like murine cell line (ATCC Number TIB-71; American Type Culture Collection, US) was purchased from the Cell Center, Peking Union Medical College. Cells were cultured at 37 °C and 5% CO 2 humidified atmosphere in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1 unit/ml penicillin, 100 mg/ml streptomycin. Cultured RAW264.7 cells were used for the transwell assay, confocal microscopy and patch clamp study.
Plasmids encoding mCFP-Orai1 and STIM1-EYFP were constructed according to the method described by Zeng et al. 43 . T-REx HEK293 cells (Invitrogen) were routinely cultured and transfected with the above two plasmids to obtain stable expression of Orai1 and STIM1 in this cell line. Transfected HEK293 cells were grown in a DMEM complete culture medium with 200 ng/ml zeocin and 1 μg/ml tetracycline to induce Orai1 expression.
Transwell assay. Transwell assays were performed to evaluate the effect of c-MWCNTs on the migration capacity of RAW264.7 cells. The transwell apparatus (Millipore Co., USA) is composed of an upper chamber (Millipore insert) and a lower chamber. The bottom of the upper chamber is a polyethylene terephthalate (PET) membrane with 8-μm pores that allow cells to penetrate/migrate. The lower chamber is a well of a 24-well culture plate. Cultured RAW264.7 cells were collected and suspended in DMEM medium, and then were counted and adjusted to a concentration of approximately 10 5 cells/ml DMEM. Cell suspensions (200 μl) were seeded onto the upper chamber and cultured in DMED with 10% FBS at 37 °C for 24 h to allow cells to fully grow and adhere to the upper face of the PET membrane so that they would be ready to migrate. The lower chamber was filled with 700 μl DMEM. c-MWCNTs or other agents (such as MCP-1, U-73122, etc.) were delivered into the DMEM solution contained in the lower chamber. Cells that migrated across the pores of the PET membrane would adhere to the lower face of the PET membrane of the Millipore insert. Note that although c-MWCNTs were placed in the lower chamber, the 8-μm pores of the PET membrane equipped in the bottom of the Millipore insert also allowed c-MWCNTs to penetrate the pores and diffuse from the lower chamber to the upper chamber driven by the c-MWCNT gradient between the two chambers, therefore providing a chance for RAW264.7 cells in the upper chamber to be exposed to c-MWCNTs and thus trigger cell migration. After the 24-h migration, the Millipore insert was removed and the DMEM medium was discarded. The insert (including the PET membrane) was washed with PBS three times, and cells adhering to the lower face of the PET membrane (indicating the migrated cells) were fixed with 4% paraformaldehyde for 15 min, and then were stained with 0.1% crystal violet for 20 min, followed by a PBS wash to remove excess stain. Cells on the upper face of the PET membrane (cells not migrated) were removed by a gentle wipe with a cotton swab. Cells adhering to the lower face of PET membrane (cells migrated) were counted under a microscope (200× objective). Five fields were randomly chosen and cells were captured and counted with Image-pro Plus 6.0 with the amount of normal control cells standardized to 1. The same experiments were performed in triplicate in three wells.

Transmission electron microscopy (TEM)
TEM was performed to evaluate the phagocytosis of RAW264.7 macrophages on c-MWCNTs according to our recent report 44 with modifications. Briefly, after the transwell experiments, RAW264.7 cells were collected from the lower face of the PET membrane of the insert (representing the migrated cells). The collected RAW264.7 cells in DMEM were centrifuged at 1000 rpm for 4 min. The cell pellets were fixed with 2.5% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.2) and stored at 4 °C until TEM was performed. To perform the TEM, RAW264.7 cells were washed with phosphate buffer and were further fixed with 1% phosphate buffered osmium tetroxide, rewashed with phosphate buffer, dehydrated by graded ethanol and acetone, infiltrated with epoxy resin and acetone, and finally were embedded in epoxy resin. Ultra-thin sections were cut with an RMC MT-X ultrathin microtome, mounted on copper grids and then stained with uranyl acetate and lead citrate and washed with distilled water. Ultrastructural images of RAW264.7 cells were shot with a GATAN digital camera (Gatan, USA) equipped in a transmission electron microscope (JEOL-1011, Japan). RAW264.7 cells without exposure to c-MWCNTs served as the control.
Patch clamp. Whole-cell patch clamp experiments were performed to record the CRAC currents (I CRAC ) of cultured RAW264.7 cells and HEK293 cells with stable expression of CRAC channels (Oria1 and STIM1 subunits) at room temperature using an EPC-10 amplifier (HEKA, Germany). Pipette resistance was controlled to 2-4 MΩ. To record I CRAC , a 200-ms step pulse was delivered from a holding potential of 0 mV to −120 mV. The bath (external) solution consisted of (in mmol/L): NaCl 155, KCl 4.5, CaCl 2 22, MgCl 2 1, D-glucose 10, HEPES 5 (pH 7.4 with NaOH). The pipette (internal) solution consisted of (in mmol/L): Cs aspartate 140, MgCl 2 3.01, CaCl 2 0.66, EGTA 11.68 (free [Ca 2+ ] i was approximately 10 nM), HEPES 10 (pH 7.2 with CsOH). Current data were analyzed using Clampfit 10.1 software, and current tracing generation was performed by OriginPro8.0 software. The c-MWCNTs were added to the pipette solution to obtain the desired concentration (50 μg/ml), thus yielding an intracellular application of c-MWCNTs to observe the "immediate" effect of c-MWCNTs on I CRAC , as the effect of extracellular c-MWCNTs on I CRAC emerged very slowly (>30 min).
Confocal microscopy was also used to evaluate the calcium mobilizing effect of c-MWCNTs in RAW264.7 cells. Briefly, RAW264.7 cells were incubated with Fluo-3/AM (3 μmol/L) in the dark for 1 h, washed with PBS three times, and then placed into DMEM containing c-MWCNTs. Cell fluorescence change was continuously monitored under a confocal microscope until the Ca 2+ transient appeared.
Enzyme linked immunosorbent assay (ELISA). ELISA was performed to measure the activity of PLC and the level of IP3 in RAW264.7 macrophages in response to c-MWCNT exposure, using the respective ELISA kits (TSZ Biosci., USA) for PLC and IP3. Briefly, after exposure to c-MWCNTs (given extracellularly) for 6 h, RAW264.7 cells were washed and then were lysed. Cell lysates were collected and centrifuged at 12,000 rpm to acquire the supernatant free of MWCNTs. The absence of c-MWCNTs in the supernatant was confirmed by light microscopy and TEM. ELISA was performed to measure the activity of PLC and the level of IP3 in the cell lysate supernatant according to the manufacturer's instructions. Optical absorbing values were read at 450 nm and corrected by 570 nm. The data were analyzed using SoftMax Pro4.8 software.
Real time quantitative PCR (RT-qPCR). RT-qPCR was performed to evaluate the potential effect of c-MWCNTs on the transcription level of PLC in RAW264.7 cells. Briefly, RAW264.7 cells were cultured in DMEM containing c-MWCNTs in a 6-cm dish and harvested at a cell density of 80-90%. Total RNA was isolated from the RAW264.7 cells using an RNA extraction kit (Tiangen Co., Beijing, China). The first strand cDNA was created with 1 μg total RNA using ReverTra Ace qPCR RT Kit (Toyobo, Japan). Primers were designed using the Premier 5.0 program. PLC, sense: ATGAAATCCTTTACCCACC, anti-sense: ACAGCGACATCCAGACA. β-actin, sense: TTGTTACCAACTGGGACGACAT, anti-sense: GTGTTGAAGGTCTCAAACATGATCT. RT-qPCR was performed using Real-time PCR Master Mix (Toyobo, Japan) with the ABI-7900 Real Time PCR System (ABI, USA). Each measurement was made in triplicate and expressed relative to the housekeeping gene β-actin.

Measurement of ROS level.
To determine whether c-MWCNTs induce production of reactive oxygen species (ROS) in RAW264.7 cells, intracellular ROS was measured using an ROS assay kit (Jiancheng Biotechnology, Nanjing, China), based on the ROS-mediated conversion of nonfluorescent 2,7-dichlorofuorescin diacetate (DCFH) into fluorescent dichlorofluorescein (DCF). After incubation with different treatments in 96-well plates, RAW264.7 cells were washed with PBS and subsequently incubated with DCFH-DA (final concentration 10 μmol/L) in PBS at 37 °C for 20 min. Then, the fluorescence of the cells was measured using flow cytometry (BD, USA). Statistical analysis. The data are presented as the mean ± standard deviation (SD). Statistical analyses were performed using Student's t-test for paired data and analysis of variance followed by Newman-Keuls multiple comparison for multiple group data. P < 0.05 was considered statistically significant.