Carnosic Acid Reverses the Inhibition of ApoE4 on Cell Surface Level of ApoER2 and Reelin Signaling Pathway

Reagents and antibodies

CA (>95% purity) was obtained from Biobiopa (Yunnan, China). ApoE4 was purchased from PeproTech (Rock Hill, NJ, USA). The antibodies used were as follows: rabbit anti-ApoER2 (Cat. No: ab108208), mouse anti-ApoER2 (Cat. No: ab58216), and rabbit anti-NMDAR2B (Cat. No: ab183942) from Abcam (Cambridge, MA, USA); mouse anti-SNX17 (Cat. No: sc-166957) and mouse anti-Dab1 (Cat. No: sc-271136) from Santa Cruz Biotechnology (Santa Cruz, CA, USA); rabbit anti-CREB (Cat. No: 9197 S) /P-CREB (Cat. No: 9198 S), rabbit anti-NMDAR2A (Cat. No: 4205 S) /P-NMDAR2A (Cat. No: 4206 S), and mouse anti-phosphotyrosine (Cat. No: 9411) from Cell Signaling Technology (Beverly, MA, USA); secondary anti-mouse and anti-rabbit antibodies conjugated to Alexa Fluor 488 or 598 from Abcam (Cambridge, MA, USA).

Cell culture

PC12 cells were purchased from ATCC. The PC12 cells were cultured in DMEM (Dulbecco’s modified Eagle’s medium, DMEM, Biological Industries, Israel) with 1% penicillin-streptomycin (Thermo Fisher Scientific, MA, USA) and 10% FBS (fetal bovine serum, FBS, Biological Industries) at 37°C with 5% CO₂ and saturated humidity.

Neuronal cultures

Pregnant Sprague-Dawley rats were acquired from the animal center at Shandong University. Hippocampal neurons were extracted and cultured as previously described [17, 18]. In brief, the hippocampus was dissected from embryos at embryonic day 18 and digested with 0.25% EDTA/trypsin at 37°C for 30 min before transferring to F12/DMEM medium containing 10% fetal bovine serum. Neurobasal medium containing 2% B27 (Thermo Fisher Scientific, MA, USA), 0.5 mM glutamine (Thermo Fisher Scientific, MA, USA) and 10% penicillin-streptomycin (Thermo Fisher Scientific, MA, USA) was used for culturing neurons at 37°C with 5% CO₂ and saturated humidity.

Cell surface biotinylation

Cells were incubated with 0.3 mg/mL EZ-Link Sulfo-NHS-SS-Biotin (Thermo Fisher Scientific, MA, USA) and lysed using RIPA lysate (Beyotime Biotechnology, Jiangsu, China). Streptavidin beads (Thermo Fisher Scientific) bound to biotinylation surface proteins were incubated with the cell lysate overnight at 4°C. The streptavidin-biotinylation surface protein complex was boiled for 5 min in SDS-PAGE loading buffer (CWBiotech, Beijing, China). The protein samples were then loaded onto SDS-PAGE gels for electrophoresis and detected by western blotting with ApoER2 antibody.

Immunofluorescence analysis

The cells were fixed, permeabilized, and blocked. ApoER2 (Abcam, Cat. No: ab58216) or SNX17 antibodies was used for incubating cells at 4°C overnight, respectively. Cells were then incubated for 1 h at room temperature using Alexa Fluor 488 conjugated or Alexa Fluor 594 conjugated secondary antibody, respectively. Images were captured with a Zeiss LSM780 confocal microscope (Microstructural Platform, Shandong University). The number of ApoER2 and SNX17 aggregates was assessed by quantifying points of overlap of red and green fluorescence. Fluorescence intensity was analyzed using ImageJ software (>50 cells per data set). Specifically, a straight line was drawn onto a magnified image and the red and green fluorescence intensities on the line were measured. Fluorescence intensity values were then plotted graphically using GraphPad Prism software.

Immunofluorescence analysis of membrane surface ApoER2

A fluorescence-based antibody binding assay was used for analyzing surface receptors described elsewhere [19, 20]. Briefly, cells were fixed, blocked, and then incubated with ApoER2 antibody (Abcam, Cat. No: ab58216, corresponding to amino acids 83-171 of Human ApoER2). Goat anti-mouse IgG H&L (Alexa Fluor 594) (Abcam, Cat. No: ab150116) was used for incubating cells. Then, the cell slides were permeabilized and blocked. ApoER2 antibody was used for incubating cells and then incubated with goat anti-mouse IgG H&L (Alexa Fluor 488) (Abcam, Cat. No: ab150113). Images were captured using a Zeiss LSM780 confocal microscope. Fluorescence intensity was analyzed using Image-Pro Plus 6.0 software (>50 cells per data set). Briefly, the grey value on the axon was measured and converted to the absolute expression value.

Western blotting

The procedure for western blotting is described in detail in our previous publication [21]. Briefly, cells were lysed in RIPA lysis buffer. Protein samples were then electrophoresed, transferred, blocked, and incubated with primary antibodies. The membranes were then incubated with peroxidase-coupled anti-mouse/rabbit IgG. Finally, the membranes were visualized using ECL (Millipore). Images of membranes were obtained using BioImaging Systems (UVP Inc., Upland, CA, USA).

NMDA receptor and Dab1 phosphorylation assays

Cells were lysed using RIPA lysis buffer. The protein supernatant was then transferred to a fresh EP tube and rotated overnight with an anti-phosphotyrosine antibody (Cat. No: 9411, Cell Signaling Technology) at 4°C. Immune complexes were precipitated by incubation with 30μL Protein A/G Magnetic Beads (MCE, NJ, USA) at 4°C for 2 h. Precipitated proteins were analyzed by western blotting using anti-Dab1, NMDA2A, or NMDA2B antibodies.

Immunoprecipitation

Cells were lysed in RIPA lysis buffer. The protein supernatant was subsequently rotated overnight at 4°C with anti-ApoER2 or anti-SNX17 antibody. Immune complexes were precipitated by incubation with 30μL Protein A/G Magnetic beads at 4°C for 2 h. Precipitated proteins were analyzed by western blotting with anti-SNX17 or anti-ApoER2 antibody.

Molecular simulation docking

The Discovery Studio software was used to assay the possible binding regions of CA with ApoER2 and SNX17. The crystal structure of the ApoER2 LA12 domain (PDB accession number 5B4Y) and the SNX17 PX domain (PDB accession number 3LUI) were acquired from the PDB database. The CA molecular structure was drawn using Discover Studio 4.1 software. The Discovery Studio software was then used to simulate how the respective domains of ApoER2 (5B4Y) and SNX17 (3LUI) may interact with CA.

RNA-seq analysis

The RNAs were isolated from PC12 cells treating with CA/APOE4 and control cells using Trizol Reagent. The libraries were obtained from the Illumina TruSeq1 RNA Sample Preparation. The samples were sequenced on an Illumina HiSeq 2500 Sequencing system. The TopHat2 software was used to map to the genome assembly. The HTSeq software was used to quantify the gene-level expression. The Deseq2 software was used to obtain differential expression.

Statistical analysis

The Student’s t-test or one-way ANOVA was used to assess statistical significance. Results are presented as mean±S.D., and p < 0.05 was considered significant.

CA increases cell surface level of ApoER2 to counteract the inhibitory effect of ApoE4

In AD, the surface level of ApoER2 has better resistance to Aβ via the Reelin-ApoER2-NMDAR pathway [4, 5]. In addition, CA can also resist Aβ damage and reduce Aβ plaques in the AD mouse model [16]. Therefore, we wanted to explore the relationship between CA and ApoER2 surface expression. PC12 cells were treated with 10μM CA for 0.5 and 2 h. Although no significant difference was found in the total expression of ApoER2 (Fig. 1A, B), surface biotinylation analyses showed that the cell surface level of ApoER2 was significantly elevated after CA treatment (Fig. 1A, C).

OPEN IN VIEWER

Fig.1

Carnosic acid (CA) promotes the accumulation of ApoER2 on the cell surface. PC12 cells were incubated with CA (10μM) for 0.5 h and 2 h, and either lysed and analyzed for total cellular ApoER2 level by western blotting or subjected to surface biotinylation before lysis, followed by affinity precipitation and analysis of biotinylated proteins by western blotting for cell surface ApoER2 (A). Bands from Total ApoER2 in A were quantified by ImageJ software (B) (n = 5; NS = no significance, 0.5 h or 2h versus control; Students t-test). Bands from surface ApoER2 in A were quantified by ImageJ software (C) (n = 5; *p<0.05, 0.5 h or 2 h versus control; ^#p<0.05, 2 h versus control; Student’s t-test).

As ApoE4 has been reported to inhibit the recycling of ApoER2 to the cell surface [7], we next analyzed whether CA could reverse the suppressive effect of ApoE4. PC12 cells and hippocampal neurons were incubated for 2 h with CA (10μM), ApoE4 (3μg/mL), or a combination of both. As previously reported, ApoE4 reduced the cell surface level of ApoER2, and this effect was reversed by CA treatment in both PC12 cells and primary hippocampal neurons (Fig. 2A, B). Immunofluorescence analyses in PC12 cells and hippocampal neurons confirmed that CA reversed the reduction in cell surface ApoER2 levels induced by ApoE4 (Fig. 2C-F).

OPEN IN VIEWER

Fig.2

Carnosic acid (CA) reverses the inhibitory effect of ApoE4 on cell surface ApoER2. PC12 cells (A) or hippocampal neurons (B) were treated as indicated and were subjected to surface biotinylation. After lysis and affinity precipitation, biotinylated proteins were analyzed by western blotting for surface ApoER2. Bands from surface ApoER2 in A and B were quantified by ImageJ software (n = 4; *p<0.05, CA or ApoE4 versus control; ^#p<0.05, CA and ApoE4 versus ApoE4; Students t-test). PC12 cells (C) or hippocampal neurons (E) were incubated with CA or ApoE4 as indicated. Surface (red) and total ApoER2 (green) were detected by immunofluorescence. The surface expression of ApoER2 was mainly observed in the neuronal axon, rather than on cell bodies (E). Relative surface levels of ApoER2 to total ApoER2 were quantified and calculated (D, F) (n = 3; *p<0.05; Students t-test). Scale bars: 10μm in (C) and 20μm in (E).

CA increases cell surface level of ApoER2 by enhancing the interaction between SNX17 and ApoER2 inhibited by ApoE4

We then try to clarify how CA counteract the inhibitory effect of ApoE4 on the cell surface level of ApoER2. It has been reported that ApoE4 sequesters ApoER2 into intracellular recycling compartment [7] and SNX17 binds to endocytic ApoER2 to promotes its cell surface accumulation [9]. Based on these reports, we investigated whether ApoE4 affects the interaction of SNX17 with ApoER2. Immunoprecipitation and subsequent western blotting in PC12 cells revealed that ApoE4 inhibited the interaction between SNX17 and ApoER2 in a time-dependent manner (Fig. 3A, B). We then tested if CA may increase the cell surface level of ApoER2 by reversing the inhibitory effect of ApoE4 on the interaction. As we found CA enhanced the interaction between SNX17 and ApoER2 in PC12 cells (Fig. 3 C, D).

OPEN IN VIEWER

Fig.3

Carnosic acid (CA) and ApoE4 have opposing effects on the interaction of SNX17 with ApoER2. PC12 cells were treated with ApoE4 (3μg/mL) for 0.5 h, 1 h, and 2 h. Co-immunoprecipitation was performed using SNX17 antibody, and western blotting was used to detect the binding ability of SNX17 to ApoER2 (A). Co-immunoprecipitation was performed using ApoER2 antibody, and western blotting was tested for the binding ability of ApoER2 to SNX17 (B); PC12 cells were treated with CA (10μM) for 1 h and 2 h. Co-immunoprecipitation was performed using SNX17 antibody, and western blotting was used to detect the binding ability of SNX17 to ApoER2 (C). Co-immunoprecipitation was performed using ApoER2 antibody, and western blotting was used to detect the binding ability of ApoER2 to SNX17 (D). The above experiment was repeated three times independently.

We then treated PC12 cells with CA (10μM), ApoE4 (3μg/mL), or a combination of both. Subsequent immunoprecipitation and western blotting demonstrated that, in addition to increasing the binding of SNX17 with ApoER2, CA reversed the ApoE4-induced inhibition of the interaction between SNX17 and ApoER2 (Fig. 4A). The same result was observed in cultured primary hippocampal neurons (Fig. 4B). In addition, double staining of SNX17 and ApoER2 by immunofluorescence in PC12 cells and hippocampal neurons showed that ApoE4 suppressed co-localization of SNX17 and ApoER2, whereas CA significantly restored the co-localization (Fig. 4C-F, Supplementary Figure 1).

OPEN IN VIEWER

Fig.4

Carnosic acid (CA) promotes the interaction of SNX17 with ApoER2 and reverses the inhibitory effect of ApoE4. PC12 cells and hippocampal neurons were treated with CA (10μM; 2h) or ApoE4 (3μg/mL; 2 h) alone and together. Co-immunoprecipitation was performed using SNX17 antibody, and western blotting was used to detect the binding ability of SNX17 to ApoER2. The binding ability of SNX17 to ApoER2 was quantified by measuring the strips in PC12 cells (A) or hippocampal neurons (B) by ImageJ software (n = 5; *p<0.05, CA or ApoE4 versus control; ^# p ^< 0.05, CA and ApoE4 versus ApoE4; Students t-test). Cultured PC12 cells (C) and hippocampal neurons (E) were treated as indicated, followed by immunofluorescence staining for SNX17 (red) and ApoER2 (green). The nuclei were stained with DAPI (blue). Co-localization of SNX17 and ApoER2 (yellow) is indicated by the inserted frames in the magnified image. The average number of ApoER2 aggregates co-localizing with SNX17 in PC12 cells (D) or hippocampal neurons (F) were calculated by ImageJ software (n = 3; *p<0.05, CA, ApoE4, and CA and ApoE4 versus control; Student’s t-test). Scale bars: 10μm in the inserted frames of C and E.

As the interaction between SNX17 and ApoER2 promotes the recycling of ApoER2 and facilitates the transportation of ApoER2 back to the plasma membrane [9–11], our results indicated that CA increases cell surface level of ApoER2 by inhibiting the effect of ApoE4 on the interaction between SNX17 and ApoER2.

CA activates Reelin signaling pathway suppressed by ApoE4

The Reelin signaling pathway is mediated by cell surface ApoER2. ApoE4 has been reported to decrease the phosphorylation of the Reelin signaling downstream targets NMDAR and CREB by sequestering ApoER2 into the intracellular compartments [7]. Now that we have proved CA promotes cell surface ApoER2 at the presence of ApoE4, we next want to investigate whether CA could reverse ApoE4-induced inhibition of NMDAR and CREB phosphorylation. We first tested the effect of CA on the cytosolic adaptor protein Disabled-1 (Dab1), which is also known to be phosphorylated in response to Reelin signaling. We found that CA increased the phosphorylation levels of Dab1 (Fig. 5A, B). PC12 cells were then incubated for 2 h with CA (10μM), ApoE4 (3μg/mL), or a combination of both. Western blotting revealed that CA reversed the reduced phosphorylation of NMDAR (Fig. 5 C) and CREB(S133) (Fig. 5D) induced by ApoE4. A similar result was observed in cultured primary hippocampal neurons (Fig. 5E, F). These results indicated that CA could activate Reelin signaling pathway suppressed by ApoE4.

OPEN IN VIEWER

Fig.5

Carnosic acid (CA) increases the phosphorylation of the Reelin signaling pathway proteins NMDAR and CREB and reverses the inhibitory effect of ApoE4. PC12 (A) and Hippocampal neurons (B) were treated with CA. followed by co-immunoprecipitating with tyrosine-phosphorylated antibody and western blotting with Dab1 antibody. PC12 cells or hippocampal neurons were incubated with CA (10μM; 2 h), ApoE4 (3μg/mL; 2 h), or a combination of both. PC12 cells (C) and hippocampal neuronal (E) protein lysates were co-immunoprecipitated with tyrosine-phosphorylated antibodies, and western blotting was used to detect NMDAR phosphorylation levels using NMDAR2A and NMDAR2B antibodies (The co-immunoprecipitation experiment was repeated independently three times). The PC12 cells (D) or hippocampal neurons (F) were treated as above, and the expression of CERB and P-CERB (S133) was detected by western blotting. CERB and P-CERB bands were quantified using ImageJ software (n = 5; *p<0.05, CA, ApoE4, and CA and ApoE4 versus control; Students t-test). RT-qPCR confirmed the Gria2 expression in CA, ApoE4, and CA and ApoE4 treated PC12 cells (G) (n = 5; *p<0.05, CA, ApoE4, and CA and ApoE4 versus control; Student’s t-test).

We then further tried to identify the downstream effector of the Reelin pathway affected by CA and ApoE4. As CREB, a transcription factor regulates a variety of genes involved in Reelin that affect neuronal development and function [22], we performed transcriptome sequencing in PC12 cells. Differentially expressed genes between ApoE4 and CA treated PC12 cells were analyzed (Supplementary Figure 2). We found 45 downregulated genes in the ApoE4 group, 74 upregulated genes in the CA group compared with control (Supplementary Tables 1 and 2). Comparing these differential genes with the CREB gene database [23], we found Gria2 (glutamate ionotropic receptor AMPA type subunit 2), as a common target gene. Further real-time quantitative PCR results confirmed that CA increased Gria2 expression levels and reversed the inhibition of Gria2 expression by ApoE4 (Fig. 5G).

CA promotes neurite outgrowth and reverses the inhibitory effects of ApoE4

We have confirmed that CA increased the cell surface level of ApoER2 and enhanced the Reelin-NMDAR-CREB signaling pathway to induce Gria2 expression in the presence of ApoE4. As Gria2, a subunit of the AMPA receptor, was reported to express in the presynaptic axonal growth cones of neurons [24] and regulate neuronal growth [25]. We then explored the effects of CA on the regulation of neurite growth. Cultured hippocampal neurons were treated with ApoE4 (3μg/mL), CA (10μM), or a combination of both (Fig. 6A). Immunofluorescence analysis confirmed that CA promoted, while ApoE4 inhibited neurite growth. Furthermore, CA treatment reversed the inhibitory effect of ApoE4 on neurite growth (Fig. 6B, C).

OPEN IN VIEWER

Fig.6

Carnosic acid (CA) promotes neuronal growth and reverses the inhibitory effect of ApoE4. Hippocampal neurons were extracted and cultured for 1 day, followed by pre-treatment with CA (10μM), ApoE4 (3μg/mL), or a combination of both for 0.5 h. Following the culture of hippocampal neurons for a further 3 days (A), the cells were fixed and analyzed by immunofluorescence (B) (Scale bars: 20μm). The normalized neuronal length was quantified and calculated using ImageJ software (C) (n = 3; *p<0.05, CA, ApoE4, and CA and ApoE4 versus control; Student’s t-test).