INCB24360 Suppresses M1-like Macrophage Formation and NLRP3 Expression Whereas Increases IL-1β Secretion in RAW264.7 and BV-2

Background: Macrophages switch between different functional phenotypes under the different physiological or path-physiological stimuli in tissue. Macrophage functional phenotype heavily affects disease progression, including inammation, injury, neurodegenerative disease, and cancers. IDO-1, a druggable target, is an immunosuppressive enzyme expressed in tissue macrophages and induced by Interferon-γ (IFN-γ). How IDO-1 inhibitors affect the functional switches of macrophage are unknown. Methods: IFN-γ were used to increase IDO-1 expression and 1-Methyl-D-tryptophan (1-MT) and INCB24360 (Epacadostat) were used to inhibit IDO-1 activity in RAW264.7 and BV-2. Western blotting, immunostaining and ELISA were used to evaluate protein expressions or secretion. RT-PCR were used to assess to transcription. TMR-Dextran and Latex beads were used to test endocytic, macropinocytic and phagocytic ability of macrophage and Fiji image and IMARIS were used to analyze images. Results: Our results showed that INCB24360 preferentially suppresses the cellular size and lopodia growth and reduces uid uptake, macropinocytosis, and phagocytic ability of IFN-γ induced or non-induced RAW264.7 and BV-2 in vitro. 1-MT and INCB24360 suppress IFN-γ induced or endogenous NLRP3 expression levels but not caspase-1 in RAW264.7 and BV-2. NLRP3 reduction induced by 1-MT and INCB24360 parallel with the decrease of NLRP3 gene transcription and an increase of IL-1β secretion. Conclusions: Our data collectively showed that inhibiting IDO-1 with its inhibitors, especially INCB24360, preferentially suppresses cellular size and lopodia growth, NLRP3 expression, the endocytic, macropinocytic, and phagocytic capacity whereas increases IL-1β secretion in macrophage. Our ndings do not rule out the roles of IDO-1 in the formation of M1-like macrophages, but they cast doubt on the robust anti-inammatory effects of IDO inhibitors in the macrophage-mediated immune response.


Western Blot Analysis
The cells were lysed in ice-cold RIPA buffer supplemented with phosphatase inhibitor PMSF 30min. The supernatant's protein concentration was measured by BCA assay kit (PD-BCA-125, Biothrive, Shanghai, China). 25μg protein was loaded to 12% SDS-PAGE electrophoresis, and then transfer the protein from gel to PVDF membranes (Millipore, MA, USA). The membrane was blocked by 5% skim milk in TBST (PBS and 0.1% Tween), and then incubated (4°C overnight) with the antibodies. After nishing incubation in primary antibodies, the membrane was rinsed by 3 x 5 min with TBST, then incubated with speciesspeci c horseradish peroxidase-conjugated secondary antibodies (1:5000, Santa Cruz, Germany) for 60 min at room temperature. After 3x10 min washes with TBST, membranes were developed with supersensitive Enhanced Chemiluminescence substrate kit (Biothrive Ltd, ECL-P-100, Shanghai, China) for visualization by Tannon

ELISA Assay
The supernatant of RAW264.7 and BV2 cells were collected to detect the level of interleukin-1β (IL-1β) and interleukin-18 (IL-18) with mice enzyme-linked immunosorbent assay kit (MeiLian, Shanghai, China) according to the instructions of the provider.

Transwell Migration Assay
Cells were resuspended in serum-free DMEM, and plate 150µl (E5 /ml cell suspension into the upper chamber of the Transwell. After that, 800µl of DMEM containing 10% FBS were added to the lower chamber. After incubating of RAW264.7 for 48h or BV-2 for 24h, the upper chambers were removed, aspirated the culture solution in an upper chamber, xed the chambers with 4% PFA for 15 minutes, washed the chambers three times with PBS, and then stained the chambers in the crystalline violet solution for 30 minutes. After staining, the upper layer of cells gently swabbed with a cotton swab moistened with PBS, rinsed with PBS three times, and dried for microscopy.

Edu Assay
Edu assay (Invitrogen)was done by following the supplier protocol. After drug treatments, an Edu working solution (10mM, 1:2000, with green uorescence) was added to the medium and then incubated for another 12h. After incubation, the cells were xed by 4% PFA for 20 min, rinsed with PBS, permeabilized by 0.2% Triton X-100 for 20 min, and then added freshly-prepared Click-iT™ reaction solution and incubated for 30 minutes. After washing, the cell culturing slides were mounted to a slide for microscopy.

Imaging and Image analysis
Leica SP8 microscopy was used to scan all images. Fiji (ImageJ) software was used to count cell and morphological analyses. IMARIS9.5 was used for particle size analysis. Statistics SPSS20.0 and GraphPad were used for data analysis.

Results
INCB24360 reduces iNOS expression in RAW264.7 and BV-2 IFN-γ polarizes macrophage to M1 type, increases IDO-1 expression (20). To test the IDO-1 inhibitor's effect on the macrophage polarization, we have treated RAW264.7 and BV-2 with IFN-γ at 20ng/mL to increase IDO-1 expression and used 20μM 1-MT and 20μM INCB24360 to inhibit IDO-1 activity. After 24h treatment, we have detected CD206, an M2 marker, iNOS, an M1 marker, IDO-1 by western blotting. Consistent with previous ndings (21,22), we found that IFN-γ treatment increased IDO-1 and iNOS expression in RAW264.7 and BV-2 ( Fig.1A The cellular morphological changes were correlated with different functional status (23). Non-polarized RAW264.7 is a small and round cell with few processes (20). If polarized, RAW264.7cells form a large round pancake-like cell, representing M1, or a slimmer cell with longer processes, representing the M2 stage (20). To see whether the decrease of iNOS expression in RAW264. 7  To see the decrease of iNOS expression and reduction of lopodia and ru ed border by INCB24360 treatment also decrease the phagocytic ability in RAW264.7 and BV-2, we have administrated TMRdextran into RAW264.7 and BV-2 after treating IFN-γ, 1-MT, and INCB24360 for 24h. After 1 hour of TMRdextran treatment, we xed the cells and counter-stained by phalloidin-Alex-488 and DAPI. After scanning under microscopy with Z-stack, we have analyzed the total amount of dextran in four group cells by ImageJ. The results showed that INCB24360 treatment signi cantly reduced DTR-dextran uptake compared to the control and IFN-γ treated group in RAW264.7 but not in BV-2 ( Fig.3A, 3B, 3D, 3E).
Macropinocytosis is a way by which macrophage uptakes large amounts of extracellular uid (25). Macropinosome formation was dependent on membrane ru ing. The decrease of lopodia on the ru ed border after the INCB24360 treatment implied that INCB24360 might reduce the macropinocytosis of RAW264.7 and BV-2. The macropinosomes' size is heterogeneous; the diameter range from 0.2μm to 5μm (26). The vesicles, which are larger than 0.75µm, were de ned as apparent macropinocytic vesicles in DTR-dextran uptake assays by some researchers (27). We have analyzed the macropinocytic vesicles that were larger than 0.75µm by Imaris9.6 software and found that INCB24360 treatment decreased the number of phagocytic vesicles larger than 0.75µm in RAW264.7 and BV-2 ( Fig.3A, C, D, F).
Macrophage, especially tissue-resident macrophage, can also phagocytize foreign-derived particulates such as alum, silica et al. M1 macrophage can phagocytize foreign-derived particulates (28, 29). We also treated RAW264.7 and BV-2 with Latex beads, a spherical polymer particle, uorescence red, and 1µm, after treating with IFN-γ, 1-MT and INCB24360 for 24h. After xing with 4% PFA, we analyzed the phagocytized Latex beads in both cells and found that RAW264.7 (Data not shown) has a weaker ability to phagocytize Latex beads compared to BV-2. We observed that 1-MT and INCB24360 treatment slightly decrease the Latex beads uptake of BV-2 compared to the control, but the decreases are not statistically signi cant (Fig.3G-J).
To test if IDO-1 inhibitors block IFN-γ-driven ru ed border formation and lopodia growth, we treated RAW264.7 and BV-2 with IFN-γ for 12h, and then added 1-MT or INCB24360 for 24h, and xed the cells, and stained with phalloidin Alexa-488 and DAPI. We also scanned the cells with confocal microscopy with Z-stack, analyzed cellular perimeters, and lopodia density on the cellular border. Our data showed that INCB24360 inhibits IFN-γ induced ru ed border and lopodia formation in RAW264.7 and BV-2 ( Fig.5E, Si-Fig.5D). The measuring and counting data showed that both 1-MT and INCB24360 treatment signi cantly inhibited the increase of cellular size induced by IFN-γ in RAW264.7 ( Fig.5F) but not in BV-2 (Si- Fig.5E); INCB24360 treatment also inhibits the increase of lopodia density on the cellular membrane induced by IFN-γ (Fig.5G, Si-Fig.5F); 1-MT treatment does not restrict the increase of lopodia density on the cellular membrane induced by IFN-γ (Fig.5G, Si-Fig.5F). Our data showed that INCB24360 inhibits IDO-1 induced cellular size increase and lopodia formation. 1-MT weakly inhibits IDO-1-induced cellular size increase.
INCB24360 suppress IFN-γ induced endocytic, macropinocytic phagocytic ability IFN-γ increased the phagocytic ability of RAW264.7 and BV-2 (Fig.3A, D After scanning under microscopy with Z-stack, we have analyzed the total amount of dextran in four group cells. The results showed that INCB24360 decreased TMR-dextran uptake ability in both RAW264.7 and BV-2; 1-MT signi cantly decreased TMR-dextran uptake ability in RAW264.7 but not in BV-2 (Fig.6A, B, D, E). We have analyzed the DTR-dextran phagocytic vesicles that were larger than 0.75µm, which is re ective of macropinocytosis, by Imaris9.6 software and found that INCB24360 treatment decreased the number of phagocytic vesicles larger than 0.75µm in RAW264.7 and BV-2, and 1-MT signi cantly decreased the number of phagocytic vesicles larger than 0.75µm in RAW264.7 but not in BV-2 ( Fig.6A, C, D, F).
To test the effect of 1-MT and INCB24360 on macrophage phagocytic ability, we treated RAW264.7 and BV-2 with IFN-γ for 12h, and then added 1-MT or INCB24360 for 24h, and added Latex beads to the culturing medium to phagocytize. After xing, we analyzed the phagocytized Latex beads in the two cells and found that INCB24360 and 1-MT treatment signi cantly decreased the count of BV-2 that phagocytize Latex beads, the number of phagocytic latex beads in each IFN-γ induced BV-2 treated with INCB24360 was signi cantly lower than that in IFN-γ induced BV-2, the number of phagocytic latex beads in each IFN-γ induced BV-2 treated with 1-MT was not signi cantly lower than that in IFN-γ induced BV-2 ( Fig.6G-J). Collectively, I-MT and INCB24360, especially INCB24360, preferentially inhibit the endocytic, phagocytic, and macropinocytic ability of M1-like macrophages induced by IFN-γ.
To test if 1-MT and INCB24360 also can decrease IFN-γ induced NLRP3 expression, we also treated RAW264.7 with IFN-γ for 12h, and then added 1-MT or INCB24360 for 24h to inhibit IDO-1 activity. Our data showed that both 1-MT and INCB24360 could inhibit IFN-γ induced NLRP3 expression but not caspase-1 (Fig.7I-K). The RT-PCR data also showed that INCB24360 and 1-MT reduced IFN-γ induced NLRP3 gene transcription in both RAW264.7 and BV-2 but not caspase-1 (Fig.7L, M, Si-Fig.6H, I). We have immunostained those cells with NLRP3 and iNOS antibodies, and immuno uorescent staining results in RAW264.7 are consistent with protein blotting results ( Fig.7N- secretion, we have treated RAW264.7 with IFN-γ at 20ng/mL, 20μM 1-MT, and 20μM INCB24360. After treating for 24h, we have detected IL-1β and IL-18 secretion levels in the medium by ELISA. Our results showed that IL-1β and IL-18 levels in the culturing medium of IFN-γ, 1-MT, and INCB24360 treated RAW264.7 and BV-2 signi cantly increased compared to the control (Fig.8A, B, Si-Fig.7A, B). It implied that endogenous IDO-1 might can prevent the overactivation of macrophages.

Discussion
Millions of versatile macrophages reside in the human organs and solid tumors (1). Its phenotype switches control macrophage physiological and pathological action (2). Thus, the functional switches of macrophages dominantly affect the disease's outcome (1,2). IDO-1, an immunosuppressive enzyme that represses T cell activities in cancer and fetus and restricts the overactivation of the immune response in in ammation (42,43). Using 1-MT and INCB24360, we found that INCB24360 could restrict the cellular size and lopodia growth and reduce their endocytic, phagocytic, and macropinocytic capacity in RAW264.7 and BV-2. 1-MT and INCB24360 treatment also reduced NLRP3 expression in RAW264.7, whereas increasing the secretion of IL-1β. Our nding implied that manipulating endogenous or induced IDO-1 in macrophages with INCB24360 induced to form of a speci c macrophage with lower endocytic, phagocytic, and macropinocytic capacity and possesses lower NLRP3 transcription, whereas higher IL-1β secretion. Our study revealed the paradox roles of endogenous and induced IDO-1 in macrophage activities, including tissue-resident macrophages and monocyte-derived macrophages.
Macrophages, including tissue-resident or tumor-associated macrophages, are the most abundant phagocytic cells in murine and human tissues, which connect innate and adapt immune response (1 1-MT treatment dramatically surges IL-1β secretion in macrophages in virus in ammation (10). The enhancement of IL-1β secretion after INCB24360 treatment is not consistent with the formation of a typical M1 macrophage. IL-1β is a critical proin ammatory cytokine in the immune response (44). Our ndings suggest that endogenous or induced IDO-1 in macrophages might prevent the overactivation of macrophage-mediated immune response by limiting IL-1β secretion. Another concern is that tryptophan also is the only source of nicotinic acid (45). IDO-1 is the rst rate-limiting enzyme that converts tryptophan into kynurenines that metabolize into nicotinic acid (45). The administration of nicotinic acid and its metabolites suppress IL-1β secretion in macrophage (46-48). If INCB24360 or 1-MT treatment reduces nicotinic acid de ciency in cells, which increases IL-1β secretion, it needs to be addressed in the future.
Previous studies showed that IDO-1 is a downstream enzyme of the NLRP3 in ammasome, and IDO-1 upregulation induced by lipopolysaccharide (LPS) is diminished in the glial cells of Nlrp3 −/− mouse, including microglia (39,49). Inhibition of IDO-1 with curcumin decreased NLRP3 expression (34). We showed that both 1-MT and INCB24360 treatment reduced endogenous or IFN-γ induced NLRP3 expression and reduced expression by inhibiting NLRP3 gene transcription. However, inhibiting NLRP3 with its inhibitors also increased IDO-1 expression. These data support that NLRP3 and IDO-1 regulate each other, and the interaction might depend on external stimuli or the cellular polarization status. The failure of INCB24360 in cancer clinical trials renders us the complexity of IDO-1 roles in the immune system (50). The differences of 1-MT and INCB24360 in inhibiting IDO-1 activity might partially explain the differences between the two drugs on macrophage polarization. Treg and CD8 + T cells population changes in tumor milieu are used to interpret the effectiveness of IDO-1 inhibitors in cancer treatment (11,37). The macrophage population prevails over other immune cells in human organs and solid cancers.
However, macrophage activities are often neglected in IDO inhibitor treatment. IL-1β is a proin ammatory cytokine, enhances the in ammatory reaction. Based on the increase of IL-1β in macrophages after treating with 1-MT and INCB24360, it is worth evaluating macrophage and neutrophils in ltration after IDO-1 inhibitors treatment. In addition, we observed that inhibiting IDO-1 with 1-MT and INCB24360 in IFN-γ induced cells not decreased IL-1β secretion and preferentially increased mTORC1 activity, which is measured by the pS6K antibody. This scenario was not seen in non-1-MT and INCB24360 RAW264.7. This observation re ected that macrophage in different status reacts differently to IDO-1 inhibitors.
Collectively, our study revealed that INCB24360 signi cantly inhibits the cellular size and lopodia formation, reduces the macrophages' endocytic, macropinocytic, and phagocytic ability and iNOS, TNFand NLRP3 expression whereas upregulates IL-1ß secretion in macrophage in vitro, implicating that INCB24360 could manipulate macrophage activities and IDO-1 might have a capacity to prevent overactivation of macrophage in immune response.

Conclusion
We selected 1-MT and INCB24360 and tested the effects of 1-MT and INCB24360 on macrophage physiological immunological activities. our data revealed that INCB24360 signi cantly inhibits the cellular size and lopodia formation, reduces the macrophages' endocytic, macropinocytic, and phagocytic ability and iNOS, TNF-and NLRP3 expression whereas upregulates IL-1β secretion in macrophage in vitro, implicating that INCB24360 could manipulate macrophage activities and IDO-1 might have a capacity to prevent overactivation of macrophage in immune response. Our ndings do not rule out the roles of IDO-1 in the formation of M1-like macrophages, but they cast doubt on the robust anti-in ammatory effects of IDO inhibitors in the macrophage-mediated immune response.     as statistical data, n ≥ 15. C, F. DTR-dextran particles that are larger than 0.75μm in the control, IFN-γ, IFN-γ+1-MT and IFN-γ +INCB groups of IFN-γ induced RAW264.7(C) or BV2(F) cells. n ≥ 15. G-J.
Phagocytosis of latex beads in BV-2 treated with IFN-γ, IFN-γ+1-MT and IFN-γ +INCB for 24h(G). The phagocytic ratio (H), the number of beads of each cell (I), the number of cell of one, two, three, or more beads (J). Scale bars, 100μm. One-way ANOVA; all data are expressed as the mean ± SEM. *, P<0.05, **, P<0.01; ns, no statistical difference.