MWCNTs (MWCNT-7) were obtained from Mitsui Chemicals Co., Ltd. Tokyo, Japan; TiO₂ (rutile type, with a mean primary size of 20 nm) was provided by Japan Cosmetic Association, Tokyo, Japan; SiO₂ (with a primary size of 10-20 nm) was purchased from Sigma-Aldrich, USA; and Fullerene C60 (with a mean primary size of 1 nm) was provided by Frontier Carbon Corporation, Japan. 10 mg of the 4 types of ENMs was suspended in saline containing 0.5% (w/v) Pluronic^® F-68 (PF-68, a non-ionic detergent from Sigma-Aldrich, USA) to a final concentration of 500 μg/ml. The MWCNT-7 suspension was homogenized for 1 minute four times (total of 4 min) using a Polytron PT1600E bench-top homogenizer (Kinematica, Switzerland) at a speed of 3000 rpm. The prepared four nanomaterial suspensions were sonicated at 600W for 5 minutes 6 times with 2-minute rest intervals using a JY92-2 sonicator (Scientz Co., Ltd, Ningbo, China). To ensure the dispersion and suspension of the nanomaterials, the suspensions were further sonicated for 5 minutes 4 times before use. Characterization of the suspended 4 nanomaterials, including shape, element analysis, and size distribution are shown in Supplementary Figures S1 – S3 .

Eight weeks-old female wild-type Sprague-Dawley (SD) rats were obtained from the Animal Center of Anhui Medical University. The rats were housed in the Animal Center and received Oriental MF basal diet and water ad libitum. The animal experiment protocols were approved by the Animal Ethics Committee of Anhui Medical University.

Twenty-five female SD rats were divided in to 5 groups and about 0.5 ml of the vehicle or 500 μg/ml MWCNT-7, TiO₂, SiO2, or C60 suspensions were administered by intratracheally spraying using an intratracheal aerosolizer (series IA-1B, Penn-century, Philadelphia, USA), as previously described (23), 2 times per week for 2 weeks. The total amount of the administered nanomaterials was 4 mg per kilogram of body weight for each rat. Three days after the last spraying, the animals were sacrificed under isoflurane anesthesia, and the lung was excised, fixed in 4% paraformaldehyde solution in phosphate-buffered saline (PBS) adjusted to pH 7.3, and processed for light microscopic examination and transmission electron microscopy (TEM) or scanning electron microscopy (SEM).

Hematoxylin-eosin (HE) stained sections of the lung tissues treated with the 4 ENM suspensions were used to observe lung inflammation and localization of the nanomaterials. For transmission electron microscopic (TEM) observation of TiO₂, SiO₂ and C60, paraffin blocks were deparaffinized and small pieces of the lung tissues were embedded in epon resin and processed for nanomaterial observation using a JEM-2100 transmission electron microscope (JEOL Co. Ltd, Tokyo, Japan). Since MWCNT-7 is difficult to be cut by an electronic microtome, scanning electron microscopy (SEM) was used to observe the MWCNT-7-treated lung tissues. Briefly, the HE-stained slides of the lung tissues were immersed in xylene for 3 days to remove the cover glass, immersed in 100% ethanol for 10 min to remove the xylene, and then air-dried for 2 hours at room temperature. The slides were then coated with platinum for observation using a Model S-4700 Field Emission SEM (Hitachi High Technologies Corporation, Tokyo, Japan) at 5–10 kV.

Eight weeks-old female wild-type SD rats were placed under deep anesthesia with intraperitoneally injected sodium pentobarbital and sacrificed by exsanguination from the inferior vena cava. The lung was excised under aseptic conditions and injected with 5 ml of saline through the trachea. After gentle shaking, the fluid in the lung was collected. This was repeated 2 more times. The collected fluid was centrifuged at 1800g for 5min at 4°C, and the supernatant (BALF) was concentrated with a concentrator tube (Millipore) and stored at -80°C until use. The cell pellet was resuspended in RPMI 1640 medium containing 10% fetal bovine serum (Gibco, USA), seeded in a six-well plate and cultured at 37°C for 90 minutes. The cells were washed with PBS three times to remove red blood cells, other cells, and cell debris. The remaining adherent cells were stained with anti-CD68, a macrophage marker, to confirm their identity. Briefly, the adherent cells were fixed in 4% paraformaldehyde and treated with 0.2% Triton X-100 containing 10% fetal bovine serum (Gibco)/1% bovine serum albumin in PBS at room temperature for 15 minutes, and then incubated with rabbit anti-CD68 (1:50 dilution, Bioss, Beijing, China) overnight at 4°C, and visualized with Cy3 labelled anti-rabbit IgG (1:100 dilution, Proteintech, Wuhan, China). After washing, the cells were counter-stanning with DAPI (Sigma-Aldrich). Images were captured with a florescence microscope (ZEISS LSM880+Airyscan, Germany). As shown in Supplementary Figure S4 , more than 95% of the adherent cells were positive for CD68. About 5×10⁵-1x10⁶ alveolar macrophages per rat were isolated.

1.6 ml of the four ENM suspensions (equal to 800 μg of each ENM) were incubated with 1 ml of the 20-fold concentrated rat BALF at 37°C in a shaker at 200 rpm for 4-6 hours, and then centrifuged at 20000×g for 30 minutes to separate the nanomaterials from the supernatants. The precipitates were re-suspended in PBS, and the suspensions were centrifuged at 20000×g for 10 minutes and the supernatants were discarded. The washing steps were repeated 2 more times. The final ENM precipitates with their bound proteins were resuspended in 50 μl of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) buffer (10% glycerol, 62.5 mM Tris-HCl [pH 6.8], 2% 2-mercaptoethanol, 2% SDS), heated for 5 minutes at 95°C, and then centrifuged at 20000×g for 30 minutes to dissociate the bound proteins from the nanomaterial precipitates. After protein quantification with BCA (BestBio, Shanghai, China), 10 μl of the final suspensions was subjected for SDS-PAGE and the gels was silver-stained with a silver staining kit (BestBio). DTT was added to 20 μl of the remaining SDS-PAGE buffer-dissociated protein solution to a final concentration of 100 mM DTT and boiled 5 minutes. The samples were analyzed by Liquid Chromatography with Tandem Mass Spectrometry (LC-MS) (Shanghai Applied Protein Technology Company, Ltd, Shanghai, China).

In vitro 1×10⁶ rat primary alveolar macrophages (PAMs) were cultured at 37°C overnight in a 6-well plate in RPMI 1640 culture medium containing 10% fetal bovine serum (Gibco). The PAMs were then washed 3 times with PBS and cultured in X-VIVO™ serum-free medium (Lonza, Belgium) The cells were then treated with saline containing 0.5% (w/v) Pluronic^® F-68, 1 μg/ml MWCNT-7, TiO₂, SiO2, or C60 suspended in saline containing 0.5% (w/v) Pluronic^® F-68, or 1 μg/ml of MWCNT-7, TiO₂, SiO2, or C60, pre-incubated for 4 hours with 1 μg/ml of recombinant human SP-A, SP-D or 10% concentrated BALF, and continued culture for 12 hours. The cells were harvested for RNA isolation and quantitative polymerase chain reaction (qPCR) analysis of cytokine expression, and the culture supernatants were collected for ELISA detection of cytokines.

Total RNA from rat PAMs was extracted with Trizol (Magen, Guangzhou, China) according to the manufacturer’s instructions. 500 ng of the RNA samples was reverse transcribed using a HiScript® II RT SuperMix kit (Vazyme Biotech, Nanjing, China), and qPCR analyses of IL-1β, IL-6, and TNF-α were then performed using the AceQqPCR SYBR Green Master Mix (Vazyme). GAPDH was used as an internal reference, and the relative expression of each gene was analyzed by the 2 ^-△△CT method. The primer pairs (forward/backward) used were the following: CAGCAGCATCTCGACAAGAG/CATCATCCCACGAGTCACAG for IL-1β; AGTT- GCCTTCTTGGGACTGA/TCCAAGATCTCCCTGAGAACA for IL-6; ACTCCCAGA- AAAGCAAGCAA/CGAGCAGGAATGAGAAGAGG for TNF-α; and GACATGCCG- CCTGGAGAAAC/AGCCCAGGATGCCCTTTAGT for GAPDH. Detection of IL-1β, IL-6, and TNF-α in the supernatants of rat primary alveolar macrophage cultures was performed with rat ELISA kits (MLBio, Shanghai, China) according to the manufacturer’s instructions.

1×10⁶ rat PAMs were seeded in 6-cm culture dishes with pre-placed microscope cover glass (NEST, China) and cultured in X-VIVO™ serum-free medium (Lonza) and cultured overnight. The cells were treated with saline containing 0.5% (w/v) Pluronic^® F-68; 1 μg/ml MWCNT-7, TiO₂, SiO2, or C60 suspended in saline containing 0.5% (w/v) Pluronic^® F-68; or 1 μg/ml of MWCNT-7, TiO₂, SiO2, or C60, pre-incubated for 4 hours with 1 μg/ml of recombinant human SP-A, SP-D or 10% concentrated BALF, and continued culture for 12 hours. The cells were then fixed with 4% paraformaldehyde and stained with HE. The Cells with phagocytosed ENMs were identified using an ECLIPSE polarizing microscope (LV100NPOL, Nikon, Japan). Total cell number and the number of the cells with phagocytosed ENMs in 10 fields (40x magnification) were counted, and the percent of the cells with phagocytosed ENMs (number of cells with phagocytosed ENMs/total number of cells) was determined.

Small interfering RNAs (siRNA) were used to knock-down the expression of LDL receptor related protein 1 (LRP1), CD14, and signal regulatory protein alpha (SIRPα). 3 pairs of siRNAs for each of the genes were provided by GenePharma Co. Ltd., Shanghai, China. The pair of siRNAs with the best silencing efficacy was determined by preliminary experiments. The siRNA sequences used for further experiments were as follows: GCUAAACUCGCUCAAUCUATT/UAGAUUGAGCGAGUUUAGCTT for CD14; CCAUCAAACGGGCAUUCAUTT/AUGAAUGCCCGUUUGAUGGTT for LRP1; and GCUCUAUGUACUCGCCAAATT/UUUGGCGAGUACAUAGAGCTT for SIRPα. 1×10⁶ rat PAMs were seeded in each well of a 6-well plate and cultured at 37°C overnight. Negative control RNA or siRNAs for LRP1, CD14, and SIRPα were transfected into the cells using Lipofectamine 2000 (Thermo Fisher, USA). 6 hours later, the culture media was changed to X-VIVO™ serum-free medium (Lonza), and then treated with 1 μg/ml MWCNT-7, TiO₂, SiO2, or C60 suspended in saline containing 0.5% (w/v) PF-68; or 1 μg/ml of MWCNT-7, TiO₂, SiO2, or C60, pre-incubated for 4 hours with 1 μg/ml of recombinant human SP-A, SP-D or 10% concentrated BALF, and continued culture for 12 hours. The cells were harvested for RNA isolation, qPCR analysis of silencing efficacy and cytokine expression, and western blotting; the culture supernatants were collected for ELISA. The silencing efficacy of LRP1, CD14, and SIRPα was analyzed by qPCR, as described above, and confirmed by western blotting. The specific primers used were CCAGGAACTTTGGCTTTGCTC/ACCGATGGACAACTTTCAGG for CD14; CCAATTGTGCATTTTTGCAG/GAATCAGGGGCATAGGTGAA for LRP1; and GTGTCTGTTGCTGCTGGAGA/GCATCTTCTGGGGTGACATT for SIRPα. The expression of LRP1, CD14, and SIRPα proteins was detected by western blotting. The cells were lysed in RIPA buffer (150mM NaCl, 50mM Tris pH 7.4, 1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, and 1mM PMSF) for 5 minutes and centrifuged at 4°C 12000 rpm for 20min. After protein quantification with a BCA kit (Bestbio), aliquots of the supernatants (20 μg protein) were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a PVDF membrane (Millipore, Boston, USA). After blocking in 5% nonfat milk, the PVDF membranes were incubated with primary antibodies (anti-GAPDH from Peprotech, USA, 1:500 dilution, as an internal control; anti-CD14 from Bioss, Beijing, China, 1:500 dilution; anti-SIRPα from Cell Signaling Technology, Danvers, USA, 1:1000 dilution; anti-LRP1 from Abcam, USA, 1:10000 dilution) at 4°C overnight. After three washings, the PVDF membrane was incubated with peroxidase conjugated anti-mouse or anti-rabbit secondary antibodies (1:10,000) for 60 minutes. The protein was visualized with ECL (Thermo Fisher) detection solution in a GEL Imaging System (Tanon, Shanghai, China). The effect of LRP1, CD1, and SIRPα knockdown on IL-1β, IL-6, and TNF-α production was analyzed by qPCR and ELISA, and its effect on ENM phagocytosis was assessed by polarized light microscopy, as described above.

Statistical analysis was performed using SSPS17 software. The statistical significance was analyzed using two tailed Student’s t test. A p value of <0.05 was considered to be significant.

Four types of ENMs with different shapes, chemical compositions, sizes, and surface properties were chosen for this study. MWCNT-7 is a carbon-based nanotube with a hydrophobic surface; TiO₂ is a rod-like particle with a hydrophilic surface; SiO₂ and C60 are round nanoparticles with a hydrophilic and hydrophobic surface, respectively. They form aggregates with different sizes in suspensions. Characterization of their shape and chemical composition is shown in Supplementary Figures S1, S2 . Their size distribution of the suspensions in 0.5% PF-68 saline was comparable to that of the suspensions in 0.5% PF-68 saline supplemented with the concentrated BALF ( Supplementary Figure S3 ).

Intratracheal spraying is a simple method for delivery of agents into the lung. In our preliminary experiment, black ink administered into the lung by intratracheal spraying was delivered evenly to each lobe and found to be taken up by AMs ( Supplementary Figure S4 ). Intratracheal administration of MWCNT-7, TiO₂, SiO₂, and C60 suspensions to the rat lung induced pulmonary inflammation to different extents. Compared with 0.5% PF-68 saline (NC, Figure 1A ), MWCNT-7 elicited a strong inflammatory response, with increased number of AMs and thickening of the alveolar epithelium ( Figure 1B ); while TiO₂ and SiO₂ ( Figures 1C, D ) had little effect on the alveolar epithelium, although particle-burdened alveolar macrophages were often observed. Similar to MWCNT-7, C60 caused a strong inflammatory response ( Figure 1E ). The number of AMs per 200◊sight field in the MWCNT-7, TiO₂, SiO₂, or C60 treated lung tissues was significantly increased compared with that of the NC ( Figure 1F ). Most of the induced AMs were found with burden ENM aggregates ( Figures 1B – E ). Further observation with SEM for MWCNT-7 (since MWCNT-7 is difficult to be cut by an electronic microtome) and TEM for TiO₂, SiO₂, and C60 revealed that the ENM aggregates were in the cytoplasm of the AMs and likely to exist in the phagosomes ( Figures 1G – J correspondingly for MWCNT-7, TiO₂, SiO₂, and C60). These results indicate that uptake by AMs is a main mechanism for clearance of the deposited particles.

Previous studies have demonstrated that ENMs are bound by proteins biological fluids, forming tiers of proteins surrounding the nanomaterials (14). This protein corona affects the biological behavior of the nanomaterial, including phagocytosis by macrophages (24). To determine whether secretory fluid in the respiratory tract and alveoli influences phagocytosis by alveolar macrophages of the 4 ENMs being studied, we prepared BALF and primary alveolar macrophages (PAMs). Immunofluorescence staining indicated that more than 95% of isolated cells were positive for CD68, a macrophage marker ( Supplementary Figure S5 ). The 4 ENMs, with or without pre-incubation of BALF, were then exposed in vitro to the isolated PAMs. The number of PAMs with internalized ENMs with or without pre-incubation of BALF were compared under polarized light microscopy. The results indicated that pre-incubation with BALF increased the uptake of MWCNT-7, TiO₂, and SiO₂ by PAMs ( Figure 2 ). The percent of PAMs with particles without pre-incubation of BALF (31.7%, 22.7%, and 18.2% for MWCNT-7, TiO₂, and SiO₂, respectively) was significantly lower than with pre-incubation of BALF (61.0%, 43.3%, and 53.7% for MWCNT-7, TiO₂, and SiO₂, respectively) ( Figure 2I ). Because C60 is isotropic and cannot be specifically visualized using polarized light microscopy, therefore, phagocytosis of C60 was not quantified ( Figures 2G, H ). The results obtained from MWCNT-7, TiO₂, and SiO₂ suggest that proteins in the BALF are likely to adsorb to ENMs and promote their uptake by PAMs.

Usually, there are two ways for internalization of ENMs into a cell: clathrin-dependent endocytosis and actin-dependent phagocytosis. For determining whether the 4 ENMs were taken up via phagocytosis, PAMs were treated with cytochalasin B (an inhibitor of actin polymerization, thus inhibiting phagocytosis) and exposed with MWCNT-7, TiO₂, and SiO₂ pre-incubated with BALF. Pre-treatment of cytochalasin B significantly reduced the percentage of the PAMs with burden-particles ( Supplementary Figure S6 ), indicating that uptake of the ENMs by AMs is mainly via phagocytosis.

For understanding possible interactions of the proteins present in the BALF with the 4 ENMs, the ENMs were incubated in 20-fold concentrated BALF, precipitated by centrifugation, washed with PBS, and resuspended in SDS-PAGE buffer. The ENM associated proteins were then subjected to SDS-PAGE. SDS-PAGE and silver staining revealed that proteins with different molecular weights and abundance absorbed to the surfaces of the 4 ENMs ( Figure 3A ). Further analysis with LC-MS identified a total of 854 proteins associated with the 4 ENMs. 332 of these proteins were found to be associated with each of the 4 ENMs ( Figure 3B ). The 120 most abundant proteins associated with each of the 4 ENMs are shown in Supplementary Table S1 . Most of these proteins were serum-derived proteins commonly found in BALF, including albumin, complement proteins, immunoglobulins, and apolipoproteins. Lung/respiratory tract derived proteins were also common.

Pulmonary surfactant is a complex mixture containing 4 surfactant-associated proteins (SP-A, SP-B, SP-C and SP-D). Notably, SP-A, SP-B, and SP-D were found to be associated with all 4 ENMs ( Figure 3C ). SP-D was among the most abundant proteins associated with MWCNT-7, TiO₂, SiO₂, and C60, while SP-B was among the most abundant proteins associated with MWCNT-7, SiO₂, and C60 ( Supplementary Table S1 ).

For further confirming that SP-A and SP-D were bound proteins, the 4 ENMs were incubated with 10-fold concentrated BALF and then separated by centrifugation. Immunofluorescence showed that the large aggregates in the precipitates were stained with anti-SP-A and anti-SP-D ( Supplementary Figure S7A ) and ELISA detection revealed that concentrations of SP-A and SP-D in the supernatants were decreased by the incubation with the 4 ENMs ( Supplementary Figure S7B ), demonstrating that SP-A and SP-D were bound to the 4 ENMs.

We initially investigated the effect of SP-A, SP-B, and SP-D on macrophage activation, using TNF-α, IL-1β and IL-6, the main pro-inflammatory cytokines produced upon macrophage activation, as markers of macrophage activation. Preliminary experiments showed that SP-B-coated ENMs had little effect on the production of these cytokines (data not shown). Similarly, SP-A, SP-D, or BALF alone in the absence of the ENMs ( Supplementary Figure S8A ), or each of the 4 ENMs alone without SP-A or SP-D coating ( Supplementary Figure S8B ) did not significantly affect cytokine production as detected by qPCR. In contrast, MWCNT-7, pre-incubated with 1μg/ml human recombinant SP-A or SP-D, stimulated a significant increase in TNF-α mRNA expression of PAMs, but not IL-1β or IL-6 mRNA expression ( Figure 4A ). Similar results were also observed when MWCNT-7 was pre-incubated with BALF ( Figure 4A ). Examination of the culture media by ELISA indicated that pre-incubation with SP-A, SP-D, and BALF enhanced MWCNT-7-induced TNF-α secretion ( Figure 4E ).

In TiO₂ treated PAMS, pre-incubation with SP-A or BALF enhanced IL-1β expression, while pre-incubation with SP-D enhanced both IL-1β and IL-6 expression ( Figure 4B ). Similarly, ELISA showed increased IL-1β secretion in the PAMs treated with TiO₂ that was pre-incubated with SP-A, SP-D, or BALF ( Figure 4F ). The response of SiO₂ treated PAMS was similar to that of TiO₂ treated PAMS: SP-A and BALF enhanced IL-1β expression and SP-D enhanced both IL-1β and IL-6 expression ( Figure 4C ). ELISA showed increased IL-1β secretion in the presence of SP-A, SP-D, and BALF ( Figure 4G ). Treatment withC60 pre-incubated with SP-A, SP-D, or BALF upregulated IL-6 expression at both the mRNA and protein secretion, as detected by qPCR and ELISA ( Figures 4H, D ).

In brief, the effect of SP-A and SP-D on macrophage activation as assessed by TNF-α, IL-1β, and IL-6 expression was similar to that of BALF. However, the specific effect on cytokine expression depended on the ENM to which the PAMs were exposed.

It has been reported that LRP1, CD14, and SIRPα are potential receptors for SP-A and SP-D (25–27). Thus, we used siRNA to knock-down the expression of these receptors ( Supplementary Figures S9A, B ). Knockdown of LRP1, CD14, or SIRPα did not significantly change mRNA expression of the cytokines in PAMs ( Figure S10A ). Similarly, neither addition of SP-A or SP-D into the culture medium, nor stimulation with each of the ENMs, affected the cytokine expression at the mRNA level in the receptors-knocked down PAMs ( Supplementary Figures S10B, C ). However, Knockdown of CD14 significantly reduced SP-A-enhanced TNF-α expression in MWCNT-7 stimulated PAMs ( Figures 5A, E ), while SP-D-enhanced TNF-α expression was increased by CD14 silencing ( Figure 5A ). TNF-α expression was also elevated by knockdown of LRP1 and SIRPα ( Figure 5A ). The marked increase in SP-A and SP-D enhanced TNF-α expression by knockdown of LRP1 suggests that LRP1 may be inhibitory to TNF-α production by PAMs.

Knockdown of CD14 decreased SP-A- or SP-D-enhanced IL-1β expression in TiO₂- and SiO₂-stimulated PAMs ( Figures 5B, C , F, G ). Silencing of CD14 expression also decreased SP-A-enhanced IL-6 expression by C60 stimulated PAMs ( Figures 5D, H ). Knockdown of LRP1 also markedly decreased IL-6 mRNA expression in C60 stimulated PAMs ( Figure 5D ). However, in contrast to mRNA expression, CD14 and LRP1 knockdown increased IL-6 protein secretion by C60 stimulated PAMS. Overall, knockdown of CD14 consistently downregulated SP-A enhanced cytokine mRNA expression, suggesting a common SP-A-CD14 axis for ENM-induced cytokine production.

Finally, we determined the effect of CD14 knockdown on the phagocytosis of ENMs by PAMs. CD14 knockdown reduced phagocytosis of MWCNT-7 pre-incubated with SP-A by PAMs: With pre-incubation of SP-A, 65.3% of the observed PAMS contained MWCNT-7 fibers while only 12.9% of CD14 knockdown PAMs contained MWCNT-7 fibers ( Figures 6A, B, K ). CD14 knockdown also reduced phagocytosis of TiO₂ and SiO₂: In the presence of SP-A, 47.3% of the PAMs contained TiO₂ particles while only 19.5% of CD14 knockdown PAMs contained TiO₂ particles ( Figures 6C, D, K ); in the presence of SP-D, 66.1% of PAMs contained TiO₂ particles while only 16.6% of CD14 knockdown PAMs contained TiO₂ particles ( Figures 6E, F, K ); in the presence of SP-A, 46.0% of the PAMs contained SiO₂ particles while only 16.7% of CD14 knockdown PAMs contained TiO₂ particles ( Figures 6G, H, K ); in the presence of SP-D, 55.5% of PAMs contained SiO₂ particles while only 11.2% of CD14 knockdown PAMs contained TiO₂ particles ( Figures 6I, J, K ). As noted above, because C60 is isotropic, phagocytosis of C60 could not be evaluated using polarized light microscopy.