Notch signaling increases PPARγ protein stability and enhances lipid uptake through AKT in IL-4-stimulated THP-1 and primary human macrophages.

Notch signaling and nuclear receptor PPARγ are involved in macrophage polarization, but cross talk between them has not been reported in macrophages. In this study, the effect of Notch signaling on PPARγ in IL‐4‐stimulated human macrophages (M(IL‐4)) was investigated using THP‐1‐derived macrophages and human monocyte‐derived macrophages as models. Human M(IL‐4) increased the expression of JAGGED1 and activated Notch signaling. Overexpression of Notch1 intracellular domain (NIC1) increased PPARγ expression, while inhibiting Notch signaling decreased PPARγ levels in M(IL‐4). NIC1 overexpression in THP‐1‐derived macrophages increased PPARγ protein stability by delaying its proteasome‐mediated degradation, but did not affect its mRNA. Phosphorylation of AKT was enhanced in NIC1‐overexpressing cells, and a specific AKT inhibitor reduced the level of PPARγ. NIC1‐overexpressing THP‐1 cells exhibited increased CD36 levels via activation of PPARγ, resulting in enhanced intracellular lipid accumulation. In summary, this study provides evidence linking Notch signaling and PPARγ via AKT in M(IL‐4).

Notch signaling and nuclear receptor PPARc are involved in macrophage polarization, but cross talk between them has not been reported in macrophages. In this study, the effect of Notch signaling on PPARc in IL-4-stimulated human macrophages (M(IL-4)) was investigated using THP-1derived macrophages and human monocyte-derived macrophages as models. Human M(IL-4) increased the expression of JAGGED1 and activated Notch signaling. Overexpression of Notch1 intracellular domain (NIC1) increased PPARc expression, while inhibiting Notch signaling decreased PPARc levels in M(IL-4). NIC1 overexpression in THP-1-derived macrophages increased PPARc protein stability by delaying its proteasome-mediated degradation, but did not affect its mRNA. Phosphorylation of AKT was enhanced in NIC1-overexpressing cells, and a specific AKT inhibitor reduced the level of PPARc. NIC1-overexpressing THP-1 cells exhibited increased CD36 levels via activation of PPARc, resulting in enhanced intracellular lipid accumulation. In summary, this study provides evidence linking Notch signaling and PPARc via AKT in M(IL-4).
Macrophages have multifaceted functions in the immune system depending on their microenvironments and the stimuli they receive [1,2]. In an in vitro system, macrophages are polarized to two opposing phenotypes: the pro-inflammatory [such as LPS-stimulated (M(LPS)] and pro-healing [such as M(IL-4)] phenotypes. Both types play important roles in tissue homeostasis and the pathogenesis of many diseases [2]. The reversal of the pro-inflammatory to the pro-healing phenotype reduced the plaque size and resulted in a good disease prognosis for atherosclerosis [3]. However, tumor-associated macrophages showed a prohealing-like phenotype and supported tumor progression and metastasis [4]. Consequently, macrophage activation is a double-edged sword, and it is expected that controlling its activation can be an alternative therapeutic choice for such chronic conditions [3,5].
PPARc is a ligand-dependent nuclear hormone receptor [14]. The target genes of PPARc in macrophages include CD36, LPL (lipoprotein lipase), and FABP4 (fatty acid binding protein P4) [15]. The products of these genes are necessary for metabolic regulation in macrophages [14,16]. Furthermore, PPARcdeficient macrophages exhibited impaired phagocytic activity to clear apoptotic cells in wounds, resulting in increased TNFa production [12].
PPARc degradation is regulated mainly by proteasomal degradation, which is mediated through PPARc E3 ligases [17]. MAPK/ERK-kinase 1/2 (MEK1/2) activation directly interacts with PPARc in the nucleus and supports its export to function in a transcriptional activity-independent manner or its degradation by the proteasome in the cytoplasm [18]. IFNc-induced PPARc phosphorylation at Ser112 by ERK1/2 targets it for ubiquitination and proteasomal degradation in adipocytes [18,19]. In addition, PI3K/AKT and protein kinase A are required for the transcriptional activation of PPARG, while increased PI3K activity reduced its protein activity in adipocytes [20,21].
The conserved Notch signaling pathway was reported to be a key factor operating during macrophage activation [22,23]. The inhibition of Notch signaling is achieved by using c-secretase inhibitor to suppress the cleavage of Notch receptors [24] or the overexpression of a dominant negative form of Mastermind-like protein (MAML), which functions as a scaffold protein for the formation of the Notch-CSL/RBP-Jj complex. Notch activation is required for the activation of pro-inflammatory macrophages [22,25]. Forced Notch activation in macrophages with the Notch ligand DLL4 switched macrophages to a pro-inflammatory phenotype with increased tumoricidal activity in a tumor model, suggesting a pro-inflammatory role [26]. In contrast, it was found in a breast cancer model that CSL/RBP-Jj is important for the differentiation of tumor-associated macrophages and that the deletion of Rbpj in murine macrophages resulted in defects in chitin-mediated M2 differentiation [27,28]. Therefore, the role of Notch signaling in M(IL-4) is still controversial and requires further investigation.
The cross talk between Notch signaling and PPARc has been reported. During keratinocyte differentiation, Jagged1 increases PPARc expression and inhibits the physical association between NF-jB p65, and PPARc, possibly through Notch activation. This association caused keratinocytes to undergo terminal differentiation [29]. In 3T3-L1 cells, a preadipocyte cell line, Notch1 upregulates PPARc and PPARd, which are necessary for adipocyte differentiation [30]. In this study, we uncovered the role of Notch signaling on the stability of PPARc in M(IL-4) through AKT. The impact and the mechanism of this cross talk in M(IL-4) using human monocyte-derived macrophage (HMDM) and THP-1derived macrophages as model are presented.

Cell culture and primary human macrophages and inhibitors
Ethics approval for the use of healthy donor blood was granted by the Institutional Review Board, Faculty of Medicine at Chulalongkorn University (IRB No. 055/60). All methods were performed in accordance with the relevant guidelines and regulations by Chulalongkorn University. Written informed consent for study participation was obtained before the samples were collected. The study methodologies conformed to the standards set by the Declaration of Helsinki. To generate HMDMs, CD14 + monocytes were separated from peripheral blood mononuclear cells by human CD14 MicroBeads (MACS Miltenyi Biotec, Bergisch Gladbach, Germany). CD14 + monocytes were maintained in complete medium [iMDM media supplemented with 5% human serum and antibiotics (HyClone

Real-time quantitative PCR
Cells were treated as indicated, and the total RNA was extracted using TRIzol Ò (Invitrogen, Waltham, MA, USA). Quantitative PCR (qPCR) was performed using 0.1-1 µg of RNA and the iQ TM SYBR Ò Green SuperMix (Bio-Rad, Hercules, CA, USA) following the manufacturer's protocol, and qPCR was performed using Bio-Rad CFX Connect Real Time System (Bio-Rad). The relative mRNA expression levels were calculated and analyzed as previously described [31].

mRNA stability assay
Cells were stimulated with IL-4 (20 ngÁmL À1 ) for 3 h prior to actinomycin D (ActD, 1 lgÁmL À1 ) (Merck, Darmstadt, Germany) treatment. Total RNA was collected at the indicated times and subjected to qPCR. The mRNA half-life was calculated using the equation: T1/2 = ln2/k, where k is the constant value of mRNA degradation [32].

Lipid staining
Cells were stimulated as indicated, and fixed cells were stained with Oil Red O solution for 10 min. Excessive dye was washed off with water four times. Cells were visualized under an inverted microscope (Olympus, Olympus Corporation, Japan), the staining was quantified by lipid elusion with 100% isopropanol, and the OD at 492 nm was measured.

Statistical analysis
All assays were performed in at least three independent experiments; for example, each experiment was repeated at least three times when using cell line, and each experiment was done using at least duplicate samples. All statistical analyses were performed using GRAPHPAD PRISM software (San Diego, CA, USA); the statistical significance was determined using two-way ANOVA, one-way ANOVA, or unpaired t-test. A P-value of less than 0.05 was considered significant.

IL-4 activates Notch signaling via JAGGED1 in THP-1-derived macrophages and HMDMs
To examine whether Notch signaling plays a role in M (IL-4), the activation of Notch signaling was examined by detection of cleaved Notch1 (Val1744). A rapid appearance of cleaved Notch1 within 15 min after IL-4 stimulation in the THP-1-derived macrophages was observed (Fig. 1A,B). Moreover, the mRNA level of one of the Notch target genes, HEY1, was also increased upon IL-4 stimulation (Fig. 1C). Similar results were obtained from HMDMs ( Fig. 1D-F). Pretreatment with c-secretase inhibitor (DAPT) before IL-4 stimulation completely abrogated the appearance of cleaved Notch1 in M(IL-4) (Fig. 1D).
To identify which Notch ligand(s) play a role in initiating Notch signaling upon IL-4 stimulation, we examined mRNA expression of Notch receptors and ligands by qPCR. The mRNA expression of all Notch ligands and receptors, except JAGGED1, was downregulated upon IL-4 stimulation in HMDMs and THP-1-derived macrophages ( Fig. 2A-D). NOTCH4 was undetectable in all experiments. Therefore, NOTCH1 and JAGGED1 were potentially involved in IL-4-activated Notch signaling. We tested whether JAGGED1 activates Notch signaling during IL-4 stimulation by using neutralizing antibody against JAGGED1. JAGGED1-neutralizing antibodies effectively reduced mRNA of HEY1 upon IL-4 stimulation (Fig. 2E), suggesting that JAGGED1 is the main ligand that triggers Notch signaling in M(IL-4). Together, these results suggest that IL-4 activates Notch signaling in human macrophages mainly via JAGGED1, which requires c-secretase.

Notch signaling increases PPARc expression but not its mRNA in M(IL-4)
To investigate the role Notch signaling plays in M(IL-4), the THP-1 cell line was subjected to retroviral transduction to overexpress NIC1 or the dominant negative form of MAML (DNMAML) to either increase or inhibit Notch signaling, respectively [33]. NIC1-overexpressing THP-1 cells showed increased HEY1 mRNA with or without IL-4 stimulation, while DNMAML overexpression resulted in the failure to upregulate HEY1 mRNA upon IL-4 activation (Fig. 3A). Therefore, NIC1 and DNMAML overexpression in the THP-1 cell line displayed the phenotypes of hyperactivation and hypoactivation of the Notch signaling, respectively.
Next, we investigated the effects of modulating Notch signaling on the level of PPARc in these cells. PPARc expression was increased in the IL-4-stimulated NIC1-overexpressing THP-1-derived macrophages, compared with that in the control, while DNMAML-overexpressing cells did not affect the PPARc levels (Fig. 3B,C). To exclude the possibility that NIC1 might modulate PMA signaling and caused increasing of PPARc, PPARc protein expression was directly investigated in NIC1-or DNMAML-overexpressing THP-1 without PMA treatment. Similar result with the PMA treatment was obtained as NIC1 alone is sufficient to increase the level of PPARc The densitometric intensity of cleaved Notch1 was normalized with b-actin and compared to that of the DMSO-treated control. * indicates statistically significant differences at P < 0.05 by paired t-test. The error bars represent mean AE SEM. (F) HEY1 mRNA expression was examined by qPCR. b-ACTIN was used as a housekeeping gene. The result was performed from 3 to 7 healthy donors or at five independent experiment in THP-1-derived macrophages. * indicates statistically significant differences at P < 0.05 by unpaired t-test. The error bars represent mean AE SD. (Fig. 3D,E). Consistent with this finding, the DAPTpretreated HMDMs (Fig. 4A,B) or THP-1-derived macrophages (Fig. 4C,D) decreased PPARc expression upon IL-4 treatment. Spontaneously increased cleaved Notch1 in THP-1 after 24 h was consistently observed with or without DMSO treatment. This increase may be the result of PMA-induced macrophage differentiation of THP-1 cell line. This observed phenomenon was much more delayed than IL-4-induced cleaved Notch1. When the PPARG mRNA was examined, NIC1 or DNMAML expression did not have any effect on the level of PPARG transcription in THP-1-derived macrophages (Fig. 5A). Taken together, these results indicated that Notch signaling enhanced PPARc protein expression without any effect on its mRNA.

Notch signaling regulates proteasomal degradation of PPARc
Because interfering with Notch signaling did not alter the mRNA level of PPARG, we next examined the effect of Notch signaling on PPARG mRNA stability by measuring PPARG mRNA half-life. As shown in Fig. 5B, control and NIC1-overexpressing THP-1derived macrophages exhibited a similar half-life of PPARG mRNA upon IL-4 stimulation. Therefore, Notch signaling did not regulate PPARG transcription or its mRNA stability. Previous reports found that the stability of PPARc was regulated by proteasomal degradation [17,34]. We then asked whether Notch signaling regulates the stability of PPARc by this mechanism. THP-1-derived macrophages were pretreated with MG132, a proteasome inhibitor, and subsequently stimulated with IL-4. The results clearly demonstrated that PPARc levels were increased in MG132-treated IL-4-stimulated cells, indicating the involvement of proteasomal degradation (Fig. 5C,D). In the NIC1-overexpressing THP-1-derived macrophages, treatment with MG132 did not further increase the level of PPARc (Fig. 5C,D). This result strongly indicated that NIC1 overexpression stabilized PPARc protein through the prevention of proteasomal degradation.   To confirm that NIC1 prolongs PPARc protein stability, the protein half-life was measured in the control and NIC1-overexpressing cells. The results showed that IL-4-stimulated NIC1-overexpressing THP-1 cells significantly extended the half-life of PPARc compared to that of the control (Fig. 5E)

Notch signaling cross talks with the AKT pathway to regulate PPARc protein stability
To explore how Notch signaling increases PPARc in M(IL-4), the level of IL-4Ra was measured in the IL-4-activated THP-1-derived macrophages. As shown in Fig. 6A, IL-4Ra decreased after IL-4 stimulation whether Notch signaling is activated or inhibited, confirming the previous report, and NIC1 or DNMAML overexpression did not alter the IL-4Ra expression level [35]. Next, the effect of Notch signaling on the downstream signaling of IL-4/IL-4R was investigated. DAPT-pretreated THP-1-derived macrophages or HMDMs were stimulated with IL-4, and the cell lysates were subjected to the detection of phosphorylation of STAT6 and AKT. DAPT-pretreated THP-1-derived macrophages (Fig. 6B,C) exhibited delayed AKT (Thr 308) phosphorylation while DAPT-pretreated HMDMs (Fig. 6E,F) decreased AKT (Thr308) phosphorylation. No effect of DAPT treatment was detected in IL-4-induced phosphorylation of STAT6 (Fig. 6D-G). This result suggests that Notch and AKT pathways cross talked in human M(IL-4).
To investigate the effect of NIC1 overexpression on AKT phosphorylation, NIC1-overexpressing THP-1derived macrophages were treated with IL-4 and the phosphorylation of AKT was detected. As shown in Fig. 7A-C, NIC1 overexpression significantly extended the duration of Thr308 phosphorylation while slightly increased the level of Ser473 phosphorylation. Next, the effect of specific inhibitors of AKT was tested on the level of PPARc in M(IL-4). The results showed that the PI3K/AKT inhibitor LY294002 reduced the PPARc levels in NIC1-overexpressing cells to the similar level as those of IL-4-stimulated THP-1-derived macrophages in a dose-dependent manner (Fig. 7D,E). We also detected the effect of the ERK1/2 inhibitor U0126 where it increased the PPARc levels (Fig. 8A,  B). Consistent with this finding, NIC1 overexpression also increased the level of phosphorylated ERK1/2 (Fig. 8C,D). To examine the possibility of the inhibitor effects on mRNA transcription, the effects of LY294002 and U0126 on PPARG were examined at the mRNA level. As shown in Fig. 8E, the treatment with LY294002 did not alter the level of PPARG in all conditions tested, whereas the treatment with U0126 significantly increased PPARG levels. Together, these data suggest that Notch signaling cross talked with the PI3K/AKT and ERK pathways to negatively regulate PPARG transcription (via ERK) or positively regulate PPARc protein stability (via PI3K/AKT).

Notch signaling enhances CD36 expression and lipid accumulation in THP-1 via PPARc
PPARc regulates various key features of M(IL-4) and IL-4-mediated induction both PPARc and 12/15lipoxygenase that can generate PPARc endogenous ligand, leading to PPARc activation and upregulation of PPARc target genes such as CD36 [36]. Therefore, the impact of Notch signaling was examined on CD36 expression and lipid accumulation. As expected, the flow cytometry result showed that NIC1 overexpression alone is sufficient for increasing surface expression CD36 protein level (Fig. 9A). We further tested whether NIC1-mediated upregulation of CD36 is mediated through PPARc by using PPARc-specific antagonist, T0070907. As shown in Fig. 9A, T0070907 (F) The relative densitometric intensity of pAKT/AKT and (G) pSTAT6/b-actin was determined and compared to those of DMSO-treated control at 0 min. The result is representative from five independent healthy donors or at least three independent experiments in THP-1derived macrophages. * indicates statistically significant differences at P < 0.05 by unpaired t-test. The error bars represent mean AE SD. treatment significantly decreased CD36 protein level (Fig. 9A) suggests that NIC1 increases CD36 expression in PPARc-dependent manner. In our system, we could not detect IL-4-induced CD36 upregulation in THP-1-derived macrophages during the duration tested in this study and this is consistent with previous report [37].
Next, intracellular lipid accumulation in CTRL-or NIC1-overexpressing THP-1-derived macrophages was investigated by Oil Red O staining. NIC1-overexpressing THP-1 cell line spontaneously accumulated lipids, and the level was increased upon IL-4 stimulation (Fig. 9B,C). Therefore, Notch signaling in M(IL-4) enhances the M(IL-4) biological functions of lipid accumulation. The proposed cross talk between Notch signaling, AKT/ERK, and PPARc is illustrated in Fig. 9D (see Discussion).

Discussion
In this study, we uncovered that Notch signaling increases the stability of PPARc and thus its biological functions in M(IL-4). This result is consistent with the reported involvement of Notch signaling in pro-healing macrophages [27,28]. This finding argues against the current model that Notch signaling is pro-inflammatory in macrophages such as in M(LPS) [22,25]. LPSstimulated NIC1-overexpressing THP-1 cell line did not significantly increased TNFa expression compared to that of the control THP-1 cell line (data not shown). It indicates that Notch signaling in macrophages may not function as an M1/M2 determining factor, but may rather operate in a context-dependent manner so that it can enforce the primary stimulating signaling in macrophages, being either pro-inflammatory (LPS) or pro-healing (IL-4).
The Notch receptor/ligand profiles in M(IL-4) of both HMDMs and THP-derived macrophages revealed that JAGGED1/2 are predominantly upregulated upon IL-4 stimulation. Indeed, blocking JAGGED1 reduced the upregulation of one of target genes of the Notch signaling HEY1, indicating that JAGGED1 is the main ligand for Notch activation in M(IL-4). We could not exclude the possibility of the involvement of other ligands in IL-4-stimulated human macrophages. Previously, Foldi et al. [38] reported that TLR stimulation induced Jagged1 upregulation in murine macrophages. How does Notch regulate PPARc stability in M(IL-4)? We reported here that by cross talk with the PI3K/ AKT pathway, Notch signaling prevents the proteasomal degradation of PPARc without affecting its mRNA transcriptional level or its stability. Similar observation was reported on the role of STAT6 in facilitating PPARc functions in IL-4-stimulated bone marrow-derived macrophages without an impact on transcription or stability of its mRNA [39]. Notch signaling has been reported to closely cross-regulate with the PI3K/AKT pathway in cancers, especially in T-cell acute lymphoblastic leukemia (T-ALL) [40].
The protein stability of PPARc is regulated mainly by proteasomes in adipocytes, and several E3 ubiquitin ligases are reported but proteasomal regulations of PPARc have not been reported in macrophages [41]. In adipocytes, E3 ubiquitin ligases function to either stabilize or facilitate proteasomal degradation of PPARc. Seven in absentia homolog 2 (SIAH2) [42]and makorin ring finger protein 1 (MKRN1) [43] facilitate PPARc degradation, while the tripartite motif protein 23 (TRIM23) regulates PPARc ubiquitination to stabilize it [44]. Furthermore, an E3 ubiquitin ligase NEDD4 is reported to stabilize PPARc by preventing it from proteasomal degradation in 3T3-L1 cells [41]. We have preliminarily examined the expression of these genes in our system but unable to validate that these E3 ubiquitin ligases play any role in M(IL-4). The identification of E3 ubiquitin ligase that are under regulation of Notch/AKT needs further investigation that will shed the light on how Notch and AKT cross talk that affects PPARc stability in M(IL-4).
As depicted in Fig. 9D, we propose that in M(IL-4), the activation of Notch signaling results in PI3K/AKT activation and prevents the proteasome-dependent degradation of PPARc. This model proposes that Notch signaling is a facilitator of IL-4-induced polarization of macrophages.
In contrast, Notch signaling alone is sufficient to activate the ERK1/2 pathway that negatively regulates The relative densitometric intensity of PPARc was normalized with b-actin and compared to that of IL-4-stimulated control. (C) Control or NIC1-overexpressing THP-1-derived macrophages were stimulated with IL-4 (20 ngÁmL À1 ) for the indicated times. Phosphorylated ERK1/2 and total ERK1/2 were examined by western blotting. (D) The relative densitometric intensity of pERK2/ERK2 compared to that of unstimulated control at 0 min. * indicates statistically significant differences at P < 0.05 by ttest. The error bars represent mean AE SD. (E) Control or NIC1-overexpressing THP-1-derived macrophages were pretreated with vehicle control DMSO, LY294002 (10 lM), or U0126 (10 lM) or for 1 h before stimulation with IL-4 (20 ngÁmL À1 ) for 3 h. PPARG mRNA expression was determined by qPCR. The result was performed in three independent experiments. a, b, and c indicated statistically significant differences when compared with unstimulated condition, DMSO-unstimulated condition, and IL-4-stimulated control (DMSO), respectively, with P < 0.05. The error bars represent mean AE SEM.
the transcription of PPARG. Notch1 activation has been reported to increase ERK1/2 expression in HER2 + breast cancer by suppressing PTEN expression [45]. In T-ALL, the Notch target gene HES1 negatively regulates PTEN expression, which is a tumor suppressor and negative regulator of the PI3K/AKT pathway [46]. However, in our study, NIC1 overexpression in THP-1 did not alter PTEN protein expression, indicating that PTEN is not involved in the cross talk with AKT/ERK1/2 (data not shown). Therefore, Notch signaling in M(IL-4) functions in dual modes in both negative and positive ways to regulate the PPARc levels, possibly for the optimal levels of PPARc protein, and perturbing this balance may lead to the dysregulated expression of PPARc and its regulatory gene network.
CD36 is a transmembrane receptor for oxidized lowdensity lipoprotein which functions in macrophage/ foam cell formation. CD36 expression is transcriptionally regulated by PPARc. [47]. IL-4 stimulation of macrophages results in PPARc expression and, as a result, increases lipid accumulation via CD36. Our results by NIC1 overexpression confirmed the function of PPARc in lipid accumulation. We have performed efferocytosis using apoptotic thymocytes in DAPTtreated and control THP-1 cell line but did not observe any difference in this function (data not shown). Apoptotic cell alone was reported to increase efferocytic activity via the induction of PPARc and CD36 in bleomycin induced lung injury mouse model [36]. Therefore, the regulation of Notch signaling to PPARc and lipid accumulation may require additional signal(s) besides IL-4 and CD36 for this biological function of M(IL-4).
Both PPARc and Notch play important roles in regulating cellular metabolisms in other cell types, including macrophages [48]. Whether they also regulate M (IL-4) metabolism warrants further investigation as recent evidences suggests that M(IL-4) utilizes a unique metabolic pathway that differs from that of M (LPS) [49].
In conclusion, Notch signaling plays both negative and positive roles in regulating mRNA and protein of PPARc through the ERK and AKT pathways in human macrophages activated by IL-4. Manipulating macrophage functions during polarization may need to take the Notch signaling into consideration for optimal outcome.