Notch Signaling Modulates Macrophage Polarization and Phagocytosis Through Direct Suppression of Signal Regulatory Protein α Expression

The Notch pathway plays critical roles in the development and functional modulation of myeloid cells. Previous studies have demonstrated that Notch activation promotes M1 polarization and phagocytosis of macrophages; however, the downstream molecular mechanisms mediating Notch signal remain elusive. In an attempt to identify Notch downstream targets in bone marrow-derived macrophages (BMDMs) using mass spectrometry, the signal regulatory protein α (SIRPα) appeared to respond to knockout of recombination signal-binding protein Jk (RBP-J), the critical transcription factor of Notch pathway, in macrophages. In this study, we validated that Notch activation could repress SIRPα expression likely via the Hes family co-repressors. SIRPα promoted macrophage M2 polarization, which was dependent on the interaction with CD47 and mediated by intracellular signaling through SHP-1. We provided evidence that Notch signal regulated macrophage polarization at least partially through SIRPα. Interestingly, Notch signal regulated macrophage phagocytosis of tumor cells through SIRPα but in a SHP-1-independent way. To access the translational value of our findings, we expressed the extracellular domains of the mouse SIRPα (mSIRPαext) to block the interaction between CD47 and SIRPα. We demonstrated that the soluble mSIRPαext polypeptides could promote M1 polarization and increase phagocytosis of tumor cells by macrophages. Taken together, our results provided new insights into the molecular mechanisms of notch-mediated macrophage polarization and further validated SIRPα as a target for tumor therapy through modulating macrophage polarization and phagocytosis.

The Notch pathway plays critical roles in the development and functional modulation of myeloid cells. Previous studies have demonstrated that Notch activation promotes M1 polarization and phagocytosis of macrophages; however, the downstream molecular mechanisms mediating Notch signal remain elusive. In an attempt to identify Notch downstream targets in bone marrow-derived macrophages (BMDMs) using mass spectrometry, the signal regulatory protein α (SIRPα) appeared to respond to knockout of recombination signal-binding protein Jk (RBP-J), the critical transcription factor of Notch pathway, in macrophages. In this study, we validated that Notch activation could repress SIRPα expression likely via the Hes family co-repressors. SIRPα promoted macrophage M2 polarization, which was dependent on the interaction with CD47 and mediated by intracellular signaling through SHP-1. We provided evidence that Notch signal regulated macrophage polarization at least partially through SIRPα. Interestingly, Notch signal regulated macrophage phagocytosis of tumor cells through SIRPα but in a SHP-1independent way. To access the translational value of our findings, we expressed the extracellular domains of the mouse SIRPα (mSIRPα ext ) to block the interaction between CD47 and SIRPα. We demonstrated that the soluble mSIRPα ext polypeptides could promote M1 polarization and increase phagocytosis of tumor cells by macrophages. Taken together, our results provided new insights into the molecular mechanisms of notch-mediated macrophage polarization and further validated SIRPα as a target for tumor therapy through modulating macrophage polarization and phagocytosis.
Keywords: macrophages, notch, signal regulatory protein α, shP-1, polarization, phagocytosis inTrODUcTiOn The important roles of microenvironmental elements in tumor development and progression have attracted intensive research attentions recently (1). Macrophages are the predominant population of tumor-infiltrating immune cells and participate in immune regulation in a polarized manner in response to microenvironmental stimuli (2). Classically activated macrophages or M1-polarized macrophages are typically induced by bacterial lipopolysaccharide (LPS) and interferon (IFN)-γ and invoke a type 1 response through secreting cytokines such as interleukin (IL)-12 and increasing antigen presentation capacity (3)(4)(5). Moreover, M1-macrophages display a stronger phagocytic activity to "eat" abnormal cells, including dying erythrocytes and mutant cells. On the other hand, M2-polarized macrophages are prototypically activated by IL-4/IL-13 and exhibit phenotypes roughly opposite to M1 macrophages (6). Tumor-associated macrophages (TAMs) share a part of M2 macrophage properties and promote tumor angiogenesis and immune editing through multiple mechanisms (7,8). Therefore, reprogramming TAMs from M2 into the M1 activation pattern might be a potential therapy for cancers, given that the critical molecular mechanisms controlling M1/M2 polarization are unveiled.
Signal regulatory protein (SIRP) α, also known as SIRPA, SHPS-1, p84, and BIT, belongs to the immunoglobulin (Ig) superfamily. SIRPα is abundantly expressed in macrophages, dendritic cells, neutrophils, and neurons (20). Its best characterized ligand is the ubiquitously expressed "don't-eat-me" signal molecule CD47. When SIRPα binds with CD47 through the Ig-like domains in its N-terminal region, signaling through SHP-1 or SHP-2 is activated in a phosphorylation-dependent manner, followed by a panel of different downstream signaling events (21)(22)(23). Kong et al. have demonstrated that SIRPα plays a critical role in regulating innate immune activation (24). Recently, they have further found that SIRPα functions as a significant regulator of TAMs in hepatoma (25).
In this study, we explored the regulatory mechanisms of Notch signal in macrophage polarization. Our results show that Notch signal regulates macrophage polarization at least partly by inhibiting SIRPα. We found that soluble mSIRPα ext polypeptide, a recombinant extracellular fragment of the mouse SIRPα (mSIRPα ext ), could promote M1 polarization of macrophages and increase their phagocytic activity to L1210 tumor cells in a CD47dependent way. Our findings extend the molecular signaling mechanisms downstream to Notch in macrophage polarization and highlight SIRPα as a novel target of Notch-mediated macrophage polarization for tumor therapy.

MaTerials anD MeThODs animals
Mice were maintained on the C57BL/6 background in a specific pathogen-free facility. RBP-J-floxed (RBP-J f ) mice (12) or ROSA-Stop f -NIC transgenic mice (14) were mated with Lyz2-Cre (#019096, Jackson Laboratory) mice to obtain mice with myeloid-specific Notch blockade (Lyz2-Cre-RBP-J f/f , with Lyz2-Cre-RBP-J +/f as a control) or activation (Lyz2-Cre-ROSA-Stop f -NIC, with ROSA-Stop f -NIC as a control) mice, respectively. Mice were genotyped with tail DNA using polymerase chain reaction (PCR) primers listed in Table S1  cell culture and Transfection L1210 murine leukemia cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, 100 µg/mL streptomycin, and 100 U/mL penicillin. Cells were maintained in 5% CO2 to 95% air. To culture bone marrow-derived macrophages (BMDMs), mononuclear cells were isolated from tibias and femurs of C57BL/6 mice. Cells were cultured at a density of 2 × 10 6 /mL in DMEM containing 10% FBS and 25 ng/mL murine macrophage-colony stimulating factor (M-CSF) (PeproTech, Rocky Hill, NJ, USA) for 7 days. In some experiments, IFNγ (20 ng/mL, PeproTech), LPS (50 ng/mL, Sigma, St. Louis, MO, USA), or IL4 (20 ng/mL, PeproTech) was added and cultured for 24 h before further analyses. Macrophages treated with PBS, LPS + IFNγ, or IL4 were named as M PBS , M LPS , or M IL4 , according to Epelmann et al. (5). Cells were transfected siRNA with Lipofectamine LTX (Invitrogen) according to the recommended protocol. Three pairs of siRNA for each target were designed and their knockdown efficiency was determined by qRT-PCR and Western blotting. The siRNA with highest efficiency was chosen for further experiments. Lentivirus was packaged by Cyagen Biosciences with commercial service (Guangzhou, China).

Plasmid construction and recombinant Protein Purification
Genes encoding the extracellular domain of mouse SIRPα (mSIRPα ext ) or CD47 (mCD47 ext ) was amplified with PCR using a mouse embryo cDNA library as a template. The primers used were listed in Table S1 in Supplementary Material. The amplified fragments were inserted into pET32a(+) plasmid (Novagen, Darmstadt, Germany), and the expressed fusion protein Trx-mSIRPα ext (55.2 kDa) or Trx-mCD47 ext (32.5 kDa) was purified as described previously in Ref. (26). pEFBOS-NIC was as described in Ref. (14).

reverse Transcription (rT)-Pcr
Total RNA was extracted from cell samples with the Trizol reagent (Invitrogen). cDNA was prepared with a reverse transcription kit (Takara, Dalian, China) following the supplier's instruction. Quantitative real-time PCR was performed using a kit (SYBR Premix EX Taq, Takara) and the ABI Prism 7500 Real-Time PCR System in triplicates, with β-actin as an internal control. Primers are listed in Table S1 in Supplementary Material.

Western Blot
Cells were harvested and the whole cell lysates were extracted on ice with the RIPA buffer containing a protease inhibitor cocktail (Beyotime, Haimen, China). Lysates were centrifuged and the supernatants were collected. Protein concentration was determined using a BCA protein assay kit (Pierce). Aliquots of protein lysates were separated by SDS-PAGE and blotted onto polyvinylidene fluoride membrane. Membranes were blocked with bovine serum albumin solution and probed with different primary antibodies and washed, followed by HRP-conjugated secondary antibodies (Table S2 in Supplementary Material). Protein bands were visualized with chemoluminescent reagents (Pierce).

In Vitro Pull-Down assay
The recombinant Trx-mSIRPα ext protein was cleaved using thrombin (Novagen) in a buffer containing 20 mM Tris-HCl, 150 mM NaCl, and 2.5 mM CaCl2 (pH8.4) at room temperature away from light for 20 h. The ProBond™ purification system (Invitrogen) was used to purify the recombinant mSIRPα ext proteins according to manufacturer's protocols. The S-tagged mSIRPα ext was mixed with purified Trx-mCD47 ext protein (ratio 1:1), and the mixture was incubated at 4°C for 2 h. Then anti-His antibodies (Sigma) pre-coupled to the Dynabeads-protein G (Invitrogen) were added and incubated at room temperature for 30 min with rotation. After washing with PBS-0.02% Tween 20, 50 mM glycine eluent was added and incubated at room temperature for 5 min with rotation. Proteins were collected and analyzed using SDS-PAGE with 15% acrylmide, followed by Western blotting with the anti-S tag and anti-His antibodies.

Proteomic analysis
Four-plex iTRAQ-based quantitative proteomics analysis was carried out using proteins isolated from BMDMs of Lyz2-Cre-RBP-J f/f or control mice. Each sample was labeled using iTRAQ 4-plex kits (AB Sciex Inc., Foster City, CA, USA) according to the manufacturer's instructions. Samples from the control BMDMs were labeled with 114, 115 tags, and samples from the RBP-J knockout BMDMs were labeled with 116, 117 tags, respectively. After labeling, the peptide samples were mixed for further LC-MS/MS analysis. The protein expression level in each sample was quantified and the fold change between control and the RBP-J knockout BMDMs was determined.

luciferase assay
The 5′ flanking sequence (−2,615 to +123) of the murine SIRPα gene was amplified by PCR with mouse genomic DNA as a template. The fragment was inserted into pGL3-basic to generate pGL3-mSIRPα-promoter. Different truncated fragments of the 5′ flanking region, as depicted in Figure 1F, were also generated by PCR and inserted into pGL3-basic (pGL3-mS-T1, 2,3, or 4). HeLa cells (2 × 10 4 ) were transfected with different reporters, NIC overexpression plasmid, and phRL-TK using Lipofectamine 2000™ (Invitrogen). The luciferase activity was assessed 24 h later using Luminoskan Ascent (Labsystems, Helsinki, Finland) and a Dual-Luciferase Reporter Assay Kit (Promega) according to the manufacturer's protocol. All luciferase activity was normalized to the Renilla luciferase activity.

chromatin immunoprecipitation (chiP)
Chromatin immunoprecipitation assay was performed using a kit (Merck Millipore) according to the manufacturer's instructions. BMDMs were polarized with LPS + IFNγ for 24 h and fixed with formaldehyde. Cross-linked immune complexes were sonicated and precipitated with anti-Hes1 antibody. DNA was extracted from the collected samples and analyzed by PCR with the primers listed in Table S1 in Supplementary Material.

Flow cytometry
L1210 cells (3 × 10 5 ) were incubated with mSIRPα ext (10 µg/mL) for 30 min on ice in dark. After washing with PBS, cells were stained with FITC-conjugated anti-S tag antibody (Sigma), followed by FACS analysis using a Calibur™ (BD Immunocytometry Systems, San Jose, CA, USA). BMDMs (3 × 10 5 ) were cultured in the presence of PBS (M PBS ) or LPS + IFNγ (M LPS ) for 24 h, and then incubated with mSIRPα ext (10 µg/mL) at 37°C for 2 h. After washing, cells were stained with anti-Ki67 (Sigma) or anti-Annexin V (BD), followed by FACS analysis. Data were analyzed using the Flowjo™ software.
In Vitro Phagocytosis assay L1210 cells were labeled with carboxyfluorescein succinimidyl amino ester (CFSE, Dojindo Molecular Technologies, Inc.) according to the recommended protocol, and loaded onto macrophages. In some cases, L1210 cells were pre-incubated with purified recombinant proteins at the concentration of 10 µg/mL at 37°C for 2 h before coculturing with macrophages. Cells were stained with anti-F4/80, rinsed with PBS, and visualized under a fluorescence microscope (BX51, Olympus). Phagocytosis was quantified by calculating the average number of ingested L1210 cells in macrophages.

statistics
Statistical analysis was performed with the Graph Pad Prism 5 software. Student's t-test or one-way ANOVA test was used for statistical analyses. Data were expressed as means ± SD. P < 0.05 was considered statistically significant.

resUlTs sirPα Was involved in notch signal-Mediated Macrophage Polarization
In an attempt to identify Notch downstream molecules involved in macrophage activation, we employed mice with myeloidspecific Notch blockade (Lyz2-Cre-RBP-J f/f or RBP-J mcKO, with Lyz2-Cre-RBP-J +/f as a control) or activation (Lyz2-Cre-ROSA-Stop f -NIC or NIC mCA , with ROSA-Stop f -NIC as a control). In an initial proteomic analysis of RBP-J deficient and control BMDMs using mass spectrometry, we found that the protein level of SIRPα was upregulated in RBP-J-deficient BMDMs ( Figure  S1A  BMDMs were stimulated and subjected to chromatin immunoprecipitation analysis with IgG or anti-Hes1 antibody. Precipitated chromatin DNA was analyzed with qPCR (g) or PCR followed by electrophoresis (h) (n = 4). Student's t test or one-way ANOVA test was used for statistical analyses. Bars represent means ± SD; *P < 0.05; **P < 0.01; ***P < 0.001.
BMDMs, as well as in RAW264.7 cells treated in the same way ( Figure S1D in Supplementary Material). Moreover, we found that RBP-J deficiency led to upregulated SIRPα expression under different stimuli (Figures 1A,B). On contrary, constitutive Notch activation resulted in a tendency of SIRPα downregulation in macrophages (Figures 1C,D). Flow cytometry confirmed that the expression of SIRPα was upregulated in RBP-J deficiency BMDMs ( Figure 1E). These data suggested that Notch signal repressed SIRPα expression in macrophages.
There are four Hes recognition sites (−2,615, −1,630, −1,268, and −68 bp) in the murine SIRPα promoter, suggesting that Notch signal might regulate SIRPα expression through Hes family co-repressors that are downstream to Notch receptors. We, therefore, constructed reporter genes with different truncated fragments of the SIRPα promoter, which were inserted into pGL3-basic ( Figure 1F). Reporter assay in the presence of NIC overexpression showed that Notch activation could significantly repress luciferase expression driven by SIRPα promoter fragments, and the Hes-binding site at −68 bp was required for Notch activation-mediated repression of the SIRPα promoter ( Figure 1F). Indeed, a ChIP assay indicated that anti-Hes1 antibody could significantly pull-down the SIRPα promoter fragment surrounding the Hes1 binding site at −68 bp in M LPS (Figures 1G,H). These data indicated that Notch signal directly inhibited expression of SIRPα through Hes proteins.
notch signal negatively regulated shP-1 Phosphorylation Triggered by sirPα-cD47 interaction Signal regulatory protein α is enriched in myeloid cells, while its ligand CD47 is universally expressed. Interaction between CD47 and SIRPα induces SHP1/SHP2 activation through tyrosine phosphorylation in the SH2 domains (27,28), leading to inhibited macrophage phagocytosis and pro-inflammatory response. BMDMs from normal mice were transfected with SIRPα siRNA (si-SIRPα) or control oligonucleotide (NC), which were proven to inhibit SIRPα expression efficiently (Figures  (n = 4). One-way ANOVA test was used for statistical analyses. Bars represent means ± SD; * or # , P < 0.05; ** or ## , P < 0.01; ***P < 0.001.  (Figures 2A,B). On the other hand, SIRPα overexpression obviously increased SHP1 phosphorylation in macrophages ( Figure S2C in Supplementary  Material; Figures 2C,D). We then examined the effect of Notch signal on SIRPα signaling using BMDMs isolated from RBP-J mcKO and control mice and stimulated with PBS, LPS + IFNγ, or IL4. The result showed that Notch signal deficiency significantly increased SHP1 phosphorylation both in M PBS and M LPS macrophages (Figures 2E,F). When the expression of SIRPα was knocked down with siRNA, the RBP-J deficiency-induced increase of SHP1 phosphorylation was canceled (Figures 2E,F).
On the other hand, forced Notch activation by NIC overexpression reduced SHP-1 phosphorylation, which was reversed by SIRPα overexpression mediated by a lentivirus (Figures 2G,H). These results indicated that Notch signal could repress SHP-1 activation in macrophages, likely through blocking SIRPα expression.

sirPα signaling Modulated Macrophage Polarization
Signal regulatory protein α was differentially expressed in macrophages under different polarization stimuli ( Figure S1D in Supplementary Material), suggesting that SIRPα-mediated signaling might be involved in macrophage polarization. To validate this, BMDMs from normal mice were transduced with SIRPαoverexpressing lentivirus and simulated with PBS, LPS + IFNγ, or IL4. qRT-PCR was carried out to access the expression of polarization markers in BMDMs. The result showed that SIRPα overexpression downregulated the expression of M LPS markers TNF-α and IL12 and upregulated M2 markers IL10 and MR under LPS + IFNγ stimulation ( Figure 3A). Moreover, BMDMs from normal mice were transfected with SIRPα siRNA or NC oligonucleotides and simulated as above. The result showed that the M LPS markers were upregulated while the M IL4 markers were downregulated upon SIRPα knockdown ( Figure 3B). Figure 3C). Furthermore, SIRPα was overexpressed in normal BMDMs using lentivirus and transfected with SHP1 siRNA simultaneously, followed by polarization stimulation as above. qRT-PCR showed that knockdown of SHP1 abrogated the effect of SIRPα overexpression on macrophage polarization ( Figure 3D). These data suggested that the SIRPα-SHP1 signaling repressed M LPS polarization upon LPS + IFNγ stimulation.

sirPα Partially Mediated the effect of notch signal on Macrophage Polarization
To access the functional significance of SIRPα downstream to Notch signal in regulating macrophage polarization, we cultured BMDMs from RBP-J mcKO mice and treated cells with PBS, LPS + IFNγ, or IL4 in the presence or absence of SIRPα siRNA. qRT-PCR and ELISA were performed to determine the expression The expression of polarization markers (a) and the production of cytokines (B) were detected (n = 4). One-way ANOVA test was used for statistical analyses.
of polarization markers and the production of cytokines. The result showed that RBP-J knockout reduced M LPS and increased M IL4 marker expression obviously, while knockdown of SIRPα could partially cancel the effects of Notch signal deficiency ( Figure 4A). ELISA also showed that blockade of Notch signal suppressed secretion of inflammatory cytokines, such as IL12 and TNF-α, and promoted the production of anti-inflammatory cytokine IL10 (Figure 4B). Similarly, inhibiting SIRPα recovered the cytokine production modulated by RBP-J mcKO ( Figure 4B). These data suggested that SIRPα as a downstream molecule of Notch signal could at least partially mediate the effect of Notch signal on macrophage polarization.
interaction Between notch signal and sirPα in regulating Phagocytosis of Macrophages  (Figure 5A), and knockdown of SIRPα with siRNA enhanced BMDM phagocytosis ( Figure 5B). Consistently, CD47 expressed on L1210 was inhibitory to SIRPα-mediated phagocytosis, because repression of CD47 in L1210 cells significantly increased phagocytosis (Figures S3C,D in Supplementary Material; Figure 5C). Compared with untransfected L1210 cells, incubation with CD47-compromised L1210 cells promoted M LPS -like phenotype of BMDMs (Figure 5D), which was dependent on SIRPα or SHP1 expression ( Figure 5E). These results suggested that interaction of CD47 on tumor cells with SIRPα on BMDMs regulated both phagocytosis and polarized activation of macrophages. RBP-J deficiency or treatment with GSI, an inhibitor of Notch signaling, reduced phagocytosis by BMDMs significantly (Figures 5F,G), and this effect of Notch signal blockade was reversed by SIRPα siRNA (Figure 5G). These data suggested that Notch signal increased phagocytosis by macrophages likely through repressing SIRPα.

soluble extracellular Domain of Mouse sirPα could interrupt the cD47-sirPα interaction
It has been shown that blocking CD47-SIRPα interaction could facilitate phagocytosis of tumor cells and promote antitumor immune response. We have also shown that soluble human CD47 could serve as an antagonist to block CD47-SIRPα interaction (26). Expression vector of the extracellular domain of mSIRPα (mSIRPα ext ) was constructed and expressed in E. coli together with the mCD47 ext ( Figure 6A). SDS-PAGE analysis of cell lysates revealed that the mSIRPα ext and the Trx-mCD47 ext proteins were successfully expressed with predicted molecular weights of 36 and 32.5 kDa, respectively ( Figure 6B). Pull-down assay followed by Western blotting showed that the mSIRPα ext protein could be pulled down by the Trx-mCD47 ext protein ( Figure 6C) but not by Trx (not shown), indicating that mSIRPα ext could bind to mCD47 ext . To examine whether the mSIRPα ext protein could bind to L1210 cells, we incubated the mSIRPα ext fusion protein with L1210 cells, and analyzed the cells using flow cytometry after staining with a FITC-conjugated anti-S tag antibody. The (F) BMDMs from RBP-J mcKO and control mice were treated as in (A). Phagocytosis was examined under an immunofluorescence microscope (n = 5). (g) BMDMs were treated with DMSO or GSI, and PBS or LPS + IFNγ. Phagocytosis was determined as in (a) (n = 5). One-way ANOVA test was used for statistical analyses. Bars represent means ± SD; *P < 0.05; **P < 0.01. result showed that cells incubated with mSIRPα ext exhibited significantly higher fluorescence intensity as compared with the control cells (Figure 6D), suggesting that the mSIRPα ext fusion protein could bind to L1210 tumor cells most likely through the interaction with CD47.

msirPα ext Polypeptide could Promote M1 Polarization and enhance the Phagocytosis of Macrophages
To verify whether mSIRPα ext could block the interaction between CD47 and SIRPα and attenuate SIRPα signaling in macrophages, we incubated BMDMs with mSIRPα ext in vitro. This treatment did not affect the proliferation or apoptosis of macrophages ( Figures S4A,B in Supplementary Material). However, qRT-PCR showed that mSIRPα ext could upregulate the expression of M LPS markers and downregulate that of M IL4 marker IL10 ( Figure 7A). Moreover, the upregulation of M LPS and downregulation of M IL4 markers by mSIRPα ext was canceled by forced overexpression of SIRPα through lentivirus-mediated transfection ( Figure 7B). We also examined the phagocytosis of L1210 cells by macrophages in the presence of mSIRPα ext . CFSE-labeled L1210 cells were incubated with mSIRPα ext in advanced, and then cocultured with M PBS -or M LPS -stimulated BMDMs in the presence of mSIRPα ext . The result showed that mSIRPα ext could enhance the phagocytosis of macrophages, and overexpression of SIRPα abrogated this effect ( Figure 7C). These results suggested that mSIRPα ext fusion protein could block the SIRPα signaling to modulate macrophage activation and phagocytosis, likely via competitive interaction with CD47.

DiscUssiOn
The Notch signaling pathway modulates various cell fate determination events during development. Moreover, Notch signal is involved in differentiation and plasticity of hematopoietic cells under both physiological and pathological conditions (29,30). Indeed, there are four types of Notch receptors in mammals, all of which could induce RBP-J-mediated Notch signal activation. We have shown that Notch1 is the most abundantly expressed Notch receptor in BMDMs (14). Several groups have reported that Notch signal regulates myeloid development and macrophage polarization through multiple downstream molecules, such as SOCS3, cylindromatosis, interferon regulatory factor 8, miR-125a, and miR-148a (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19). Although these studies have pointed to that Notch signal was required for macrophages M1 polarization, Foldi et al. have demonstrated that Notch signal promotes M2 activation of peritoneal macrophages in an in vivo model of chitin-induced M2 polarization (31). This probably suggested that the activation modes of macrophages modulated by Notch signal are context-dependent and influenced by specific pathological processes. Therefore, it should be of significance to further identify other potential downstream molecules of Notch signal. To understand the molecular network modulating macrophage activation downstream to Notch signal, we compared protein expression profiles between wild type and RBP-J deficient BMDMs using iTRAQ and found that SIRPα expression increased in the absence of Notch signaling. qRT-PCR and Western blotting validated the conclusion that SIRPα was downregulated by Notch signal activation or M LPS stimulation in macrophages, while M IL4 stimulation or Notch blockade upregulated SIRPα. Further experiments showed that Notch activation repressed SIRPα transcription directly through Hes1-binding sites in its promoter region. Functionally, our data suggested that Notch signal modulated macrophage polarization at least partially through regulating the expression of SIRPα. Signal regulatory protein α is a myeloid-specific receptor of CD47, which is broadly expressed on many types of somatic cells including tumor cells. SIRPα binds with CD47 and delivers a "don't eat me" signal for phagocytic cells to help tumor cells escape from immune clearance (32)(33)(34)(35)(36). Being activated by signal from adjacent cells via cell contact, SIRPα could induce intracellular downstream signaling by a cascade of phosphorylation modifications. The predominant pathway of SIRPα is mediated by SHP-1 activation, which could further influence NF-κB and Akt signaling and then regulate macrophage immune suppression (37,38). Recent studies have mainly focused on CD47 expression on tumor cells and have developed cancer immunotherapy targeting CD47 (39)(40)(41), but the regulation of SIRPα expression and its role in macrophages in tumor microenvironment have been elusive. The present study suggests that SIRPα functions as an important modulator of macrophage polarization. The expression of SIRPα is significantly different in differentially polarized macrophages. Overexpression of SIRPα in BMDMs promoted M2 polarization, while SIRPα knockdown promoted M1 polarization. Interference of CD47 or SHP-1 of SIRPα signaling in BMDMs also promoted M1 polarization, suggesting that SIRPα regulates macrophage polarization dependent on interaction with CD47 and intracellular SHP-1 signaling.
Macrophages have remarkable potential as mediators of anticancer therapies based on their robust ability to ingest tumor cells and modulate tumor microenvironment. The CD47-SIRPα interaction between tumor cells and immune cells represents a critical intercellular communication that inhibits the activation of macrophage-mediated phagocytosis of tumors and thereby acts as a myeloid-specific immune checkpoint (20)(21)(22)(23). In this study, we observed that SIRPα overexpression in BMDMs decreased phagocytosis of L1210 leukemia cells, while SIRPα knockdown in BMDMs and CD47 knockdown in L1210 cells both increased phagocytosis by macrophages. Meanwhile, FigUre 7 | Recombinant mSIRPα ext promoted M1 polarization and enhanced phagocytosis of M1 macrophages in vitro. (a) Differentially polarized bone marrowderived macrophages (BMDMs) were incubated with PBS or mSIRPα ext for 6 h. The expression of TNF-α, IL12, IL10, and MR was determined by qRT-polymerase chain reaction (PCR) (n = 3). (B) BMDMs overexpressing SIRPα were incubated with mSIRPα ext for competitive interaction and stimulated with PBS or lipopolysaccharide (LPS) + IFNγ. The expression of IL12 and IL10 was determined by qRT-PCR (n = 3). (c) BMDMs were stimulated with PBS or LPS + IFNγ in the presence of mSIRPα ext . CFSE-labeled L1210 cells were then loaded and incubated for 2 h, and examined under an immunofluorescence microscope (n = 5). One-way ANOVA test was used for statistical analyses. Bars represent means ± SD, n = 3; *P < 0.05; **P < 0.01; ***P < 0.001. knock down of SHP-1, which modulates M LPS polarization in BMDMs, did not show significant effect on macrophage phagocytosis (data not shown). These findings might suggest that SIRPα-mediated polarization and phagocytosis could be mediated by different mechanisms. Molecular events downstream to these important processes need further studies. In our study, activation of Notch signaling could suppress SIRPα expression and inhibit M IL4 macrophage polarization as well as phagocytosis, which might have important implications in tumor microenvironment.
Blocking CD47 triggers the elimination of cancer cells. Nevertheless, the ubiquitous expression of CD47 on normal cells, such as red blood cells, might create a large antigen sink and unintended binding. The unwanted interaction of CD47 on normal cells could be minimized by reducing the binding strength for CD47 (42). On the other hand, high-affinity SIRPα monomers (~14 kDa) are not sufficient to induce macrophage phagocytosis, instead, act as an adjuvant to lower the threshold for phagocytosis in the presence of a separate, tumor-binding antibody. Agents that block SIRPα directly, such as anti-SIRPα antibodies, also act as adjuvants for tumor-binding antibodies (43,44). To overcome these limitations, we here validated a soluble extracellular domain of mouse SIRPα (36 KD) and confirmed that mSIRPα ext could not only promote M1 polarization but also increase phagocytosis of L1210 leukemia cells by macrophages. Therefore, mSIRPα ext might have a promising potential for the clinical application as a new therapeutic reagent. More studies are needed to assess its future pharmacological and clinical values.