Activation of the Akt1-CREB pathway promotes RNF146 expression to inhibit PARP1-mediated neuronal death

Abstract

Progressive degeneration of dopaminergic neurons characterizes Parkinson’s disease (PD). This neuronal loss occurs through diverse mechanisms, including a form of programmed cell death dependent on poly-(ADP-ribose) polymerase-1 (PARP1) called parthanatos. Deficient activity of the kinase Akt1 and aggregation of the protein α-synuclein are also implicated in disease pathogenesis. Here, we found that Akt1 suppressed parthanatos in dopaminergic neurons through a transcriptional mechanism. Overexpressing constitutively active Akt1 in SH-SY5Y cells or culturing cells with chlorogenic acid (a polyphenol found in coffee that activates Akt1) stimulated the CREB-dependent transcriptional activation of the gene encoding the E3 ubiquitin ligase RNF146. RNF146 inhibited PARP1 not through its E3 ligase function but rather by binding to and sequestering PAR, which enhanced the survival of cultured cells exposed to the dopaminergic neuronal toxin 6-OHDA or α-synuclein aggregation. In mice, intraperitoneal administration of chlorogenic acid activated the Akt1-CREB-RNF146 pathway in the brain and provided neuroprotection against both 6-OHDA and combinatorial α-synucleinopathy in an RNF146-dependent manner. Furthermore, dysregulation of the Akt1-CREB pathway was observed in postmortem brain samples from PD patients. The findings suggest that therapeutic restoration of RNF146 expression, such as by activating the Akt1-CREB pathway, might halt neurodegeneration in PD.

Introduction

Parkinson’s disease (PD) is clinically characterized by progressive motor deficits such as muscle rigidity, bradykinesia, tremor at rest, and postural instability (1). Progressive and relatively selective loss of dopaminergic neurons in the substantia nigra pars compacta (SNpc) is responsible for these clinical motor symptoms in patients with PD (1). Therefore, it is imperative to understand the molecular mechanisms of cell death pathways to develop disease modifying compounds to halt the progression of dopaminergic neuronal death. Several studies using animal and postmortem PD patient brain samples have identified several distinct cell death pathways that are involved in the loss of dopaminergic neurons (2). Apoptotic signatures have been reported in postmortem PD patient brains with morphological alterations (2); increased activation of caspase-3, caspase-8, and caspase-9 (3–5); and increased terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining (6) in the disease affected brain regions, indicating apoptotic cell death. Although apoptosis is the most extensively characterized cell death mechanism in PD, other cell death pathways may also play a role in dopaminergic neuronal degeneration (7). For example, poly (ADP-ribose) polymerase-1 (PARP1) overactivation has been seen in postmortem brain samples of patients with PD (8, 9). PARP1 is a nuclear enzyme that is overactivated on sensing oxidative-stress-induced DNA damage (8, 10, 11). This overactivation of PARP1 has been reported to cause energy depletion and mitochondrial abnormalities, ultimately leading to cell death by a distinct pathway known as parthanatos (12, 13). In animal models of PD induced by either 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) toxicity or aminoacyl tRNA synthetase complex interacting multifunctional protein 2 (AIMP2) overexpression, PARP1 overactivation and overproduction of poly (ADP-ribose) (PAR) were observed in degenerating conditions (9, 14). This type of dopaminergic neuronal death was completely reversed with pharmacological or genetic ablation of PARP1 (9, 14), indicating the significant role of PARP1 activation in the execution of dopaminergic cell loss in these models. Given the complex involvement of two distinct cell death pathways in PD, it is important to develop therapeutic strategies that could modulate both pathways to enhance therapeutic efficacy.

The role of the Ser/Thr protein kinase Akt1 has been studied most extensively in the prevention of apoptosis (15). Recruitment of Akt1 on the plasma membrane occurs through Akt1 docking on phosphatidylinositol (3,4,5) triphosphate (PIP3), which requires growth factor-mediated activation of phosphatidylinositol 3-kinase (PI3K) (15). Akt1 activation in response to oxidative stress protects cells from apoptosis (16, 17). Several molecular targets of Akt1 are involved in the mediation of anti-apoptotic function of Akt1 activation (15). Mdm2 phosphorylation by Akt1 can directly increase the E3 ligase function of Mdm2 and lead to proteasomal degradation of its target substrate p53, which plays an important role in apoptotic cell death (18, 19). Transcriptional control downstream of Akt1 activation is mediated by several transcription factors, such as forkhead box O1 (FOXO1) and cAMP response element-binding protein (CREB). Particularly, Akt1-mediated phosphorylation of CREB at Ser¹³³ increases the expression of anti-apoptotic proteins, such as Bcl-2 and Mcl-1 (20, 21). Because Akt1 activation prevents oxidative-stress-induced cell death (16, 17), which involves DNA damage and PARP1 activation, it is possible that Akt1 can also regulate PARP1-dependent cell death through a currently unknown mechanism.

The expression of E3 ubiquitin ligase RNF146 is increased by low, sublethal doses of preconditioning stimuli (meaning, prior stimuli that are necessary for altered cellular response to subsequently-given relevant stresses) and by brief exposure to glucose deprivation or hypoxia (10, 22). Expression of RNF146 enhances neuronal survival by repressing PARP1 activity in mouse models of stroke and of 6-hydroxydopamine (6-OHDA)-induced PD (8, 10). RNF146 recognizes overactivated (self-PARylated) PARP1 and subsequently promotes its polyubiquitination and proteasomal degradation (10, 11). Estrogen receptor (ER) activation can stimulate the RNF146 promoter (8), but the mechanism by which preconditioning signals activate RNF146 promoter activity is unknown.

Previously, we have identified several RNF146-inducing compounds, including the natural ER agonist liquiritigenin, which increases RNF146 expression in mouse brains and provides substantial neuroprotection against 6-OHDA–induced dopaminergic cell loss, and chlorogenic acid, which increases RNF146 promoter activity in luciferase-assay-based screens (8). Chlorogenic acid is a polyphenolic compound that is abundant in green coffee beans (23), and is protective in several animal models of neurodegenerative disease, including PD (24). Akt1 has been identified as the molecular target of chlorogenic acid (25). Chlorogenic acid directly binds to and activates Akt1 as well as its downstream targets without plasma membrane recruitment of Akt1 (25).

In this study, we further explored these mechanisms in a neuronal cell line and two PD mouse models, with supporting data from postmortem PD patient brain tissue. We found that chlorogenic acid increased RNF146 expression through an Akt1-dependent transcriptional mechanism mediated by CREB that consequently suppressed PARP1 overactivation and parthanatic (PARP1-dependent) dopaminergic cell death. The results suggest that activating the Akt1-CREB-RNF146 pathway might be therapeutically beneficial in PD patients. Additionally, the identification of a regulatory mechanism between Akt1 and parthanatos may have broader relevance.

Results

Chlorogenic acid increases RNF146 expression and prevents parthanatos

Chlorogenic acid has been previously identified to induce RNF146 expression (8). Consistent with that report, we observed increased RNF146 promoter activity in response to chlorogenic acid in SH-SY5Y cells transfected with the luciferase construct harboring a 1.9-kb region of RNF146 promoter, compared to cells treated with the vehicle control (Fig. 1A). We also observed a dose-dependent increase of RNF146 mRNA expression (Fig. 1B) and RNF146 protein abundance (Fig. 1C). Furthermore, 6-OHDA–induced cell toxicity was largely prevented by treatment with the PARP inhibitor 3AB (Fig. 1D). This result is consistent with a previous report showing a cytoprotective effect of RNF146 expression against PARP1-dependent parthanatic cell death (10), but it also suggests that parthanatos is a mechanism of 6-OHDA-induced cell toxicity. Chlorogenic acid treatment also prevented 6-OHDA-induced cell toxicity, to a similar extent to that provided by PARP inhibitor 3AB (Fig. 1D). Notably, higher doses of chlorogenic acid (30, 50, and 100 μM) treatment did not cause any cellular toxicity in SH-SY5Y cells (fig. S1A).

Fig. 1. Chlorogenic acid, an RNF146-inducing compound, is cytoprotective against 6-OHDA-induced stress.

(A) Quantification of relative RNF146 promoter activity in SH-SY5Y cells transfected with a luciferase construct of RNF146 promoter (pGL3-RNF146-Luc) and pRL-TK (for 61 hours) and subsequently treated with chlorogenic acid (10 μM for 37 hours) or DMSO as control. n = 4 separate experiments per group. (B and C) Relative abundance of RNF146 mRNA by RT-qPCR (B) and RNF146 protein by Western blotting (C) in SH-SY5Y cells treated with the indicated concentrations of chlorogenic acid for 37 hours. n = 3 separate experiments per group. (D) Trypan blue exclusion viability assay in SH-SY5Y cells pre-treated with chlorogenic acid (10 μM for 37 hours) or 3AB (10 μM for 20 hours) then treated with 6-OHDA (70 μM for 16 hours). n = 6 separate experiments per group. Data in all panels are mean ± SEM; *P < 0.05 and ***P < 0.001 by unpaired two-tailed student’s t test (A) or ANOVA test followed by Tukey’s post hoc analysis (B to D).

Cytoprotection by RNF146 expression was further examined in SH-SY5Y cells by transient transfection of a GFP-tagged RNF146 (GFP-RNF146) construct. Expression of exogenous RNF146 prevented 6-OHDA-induced cell death to a similar extent to what was exerted by 3AB and chlorogenic acid compared to mock (GFP)-transfected SH-SY5Y cells (Fig. 2A, and fig. S1B). To determine whether PARP1 activity regulated cellular viability in this experimental setting, PAR-conjugated proteins in the total protein lysates were monitored using a specific antibody—an indirect measure of PARP activation. As previously shown, 6-OHDA treatment led to a robust increase of PAR-conjugated proteins within 30 min of 6-OHDA treatment, indicating overstimulation of PARP1 by oxidative stress (Fig. 2B, C). However, this increase in PAR-conjugated proteins was suppressed by PARP inhibition with 3AB or by increased RNF146 expression—either by chlorogenic acid treatment or exogenous GFP-RNF146 transfection (Fig. 2B, C, and fig. S1C).

Fig. 2. RNF146 expression by chlorogenic acid prevents PARP1 activation and cell death induced by 6-OHDA.

(A) Trypan blue exclusion assay assessing 6-OHDA toxicity (70 μM, 16 hours) in SH-SY5Y cells treated with either chlorogenic acid (10 μM pre-treatment for 37 hours) or 3AB (10 μM pre-treatment for 20 hours) or transfected with GFP-RNF146 (for 61 hours) or a GFP mock construct control. n = 6 separate experiments per group. (B and C) Western blotting analysis of PAR-conjugated proteins and RNF146 in total lysates from cells described in (A), as indicated, using antibodies to PAR, RNF146, and GFP. n = 3 separate experiments per group. (D) Trypan blue exclusion assay in SH-SY5Y cells pretreated with chlorogenic acid (10 μM for 37 hours) and transfected with control or Rnf146-targeted shRNA (for 85 hours), with or without overexpression of mouse RNF146 (mRNF146), then treated with 6-OHDA (70 μM for 16 hours). n = 6 separate experiments per group. (E and F) Western blotting as described in (B and C) on lysates from cells described in (D). n = 3 separate experiments per group. Data are mean ± SEM; *P < 0.05, **P < 0.01, and ***P < 0.001 by ANOVA test followed by Tukey’s post hoc analysis. β-actin was used as an internal loading control in blotting assays (B and E).

Deletion of the WWE domain, but not of the catalytic ring finger (RF) domain, in RNF146 caused its failure to provide cytoprotection against 6-OHDA-induced toxicity (fig. S1D). Likewise, the 6-OHDA-induced increase in PAR-modified proteins was largely suppressed by both wild-type GFP-RNF146 and the RF-deletion mutant, but not by the WWE domain-deletion mutant (fig. S1E, F). This suggests that the WWE domain plays an essential role in RNF146’s repression of PARP activation and consequent cell death after oxidative stress.

Excess PAR synthesis by overstimulated PARP1 induces the intracellular translocation of apoptosis-inducing factor (AIF) from the mitochondria to the nuclear compartment (13, 26). Consistent with that report and with its activation of PARP1, 6-OHDA treatment induced the redistribution of AIF to the nuclear subcellular fraction (fig. S1G–J). However, either chlorogenic acid treatment or GFP-RNF146 expression—both which prevent PARP1 activation and cytotoxicity—repressed this nuclear translocation of AIF by 6-OHDA (fig. S1G–J).

Next, we sought to determine whether chlorogenic acid-mediated cytoprotection and PARP inhibition require RNF146 expression. To knockdown endogenous RNF146, SH-SY5Y cells were transfected with targeted shRNA (shRNF146). Chlorogenic acid-mediated cell survival was entirely abolished when RNF146 was knocked down by shRNA (Fig. 2, D to F). Consistent with this result, shRNA-mediated RNF146 suppression failed to block the 6-OHDA-induced increase of PAR-modified proteins (Fig. 2E, F). These results demonstrate that chlorogenic acid increases cell survival and represses 6-OHDA–induced PARP activation through RNF146 expression.

Akt1-CREB pathway activation by chlorogenic acid contributes to RNF146 induction

The promoter of RNF146 contains an ER-binding motif, and ER activation plays a role in RNF146 induction by liquiritigenin treatment (8). To determine whether chlorogenic acid also induces RNF146 expression through the ER pathway, RNF146 mRNA expression was monitored following chlorogenic acid treatment in the presence of the ER inhibitor tamoxifen. Tamoxifen treatment had no effect on chlorogenic acid-induced RNF146 mRNA expression (fig. S2A). A similar result was obtained with RNF146 protein expression using western blotting. The chlorogenic acid-induced increase of RNF146 protein expression was not influenced by pharmacological inhibition of ER (fig. S2B, C). Together, these results indicate that chlorogenic acid-induced RNF146 expression functions independently of the ER pathway.

Next, we sought to determine novel molecular mechanisms of RNF146 induction by chlorogenic acid. While chlorogenic acid has various biological effects, Akt1 has been reported to be a direct target of chlorogenic acid (24). Moreover, activation of the Akt1 pathway is critical in the regulation of cell survival (15). We monitored Akt1 activation and subsequent CREB phosphorylation by western blotting in SH-SY5Y cells treated with chlorogenic acid. Akt1 was activated by chlorogenic acid, as evidenced by an increase in Akt1 phosphorylation (Fig. 3A, B). However, Akt2 phosphorylation was not altered by chlorogenic acid treatment (Fig. 3A, B). Consistent with the fact that downstream phosphorylation of CREB by Akt1 is a well-established signaling pathway (15), we observed an approximate three-fold increase in CREB phosphorylation (Fig. 3A, B). We next examined whether Akt1 activation alone is sufficient for CREB phosphorylation and RNF146 expression. To induce Akt1 activation, SH-SY5Y cells were transiently transfected with a constitutively active form of Akt1 (Akt1-CA). The dominant negative form of Akt1 (Akt1-DN) was used as a negative control. Even without chlorogenic acid treatment, forced activation of Akt1 by Akt1-CA expression was sufficient to phosphorylate CREB and increase RNF146 expression (Fig. 3C, D). However, the Akt1-DN mutant failed to phosphorylate CREB and induce RNF146 expression (Fig. 3C, D). Pharmacological Akt1 inhibition by MK-2206 treatment further demonstrated the requirement of Akt1 activation by chlorogenic acid for cytoprotection against 6-OHDA-induced oxidative stress (fig. S2D). Pharmacological inhibition of Akt1 activation prevented the induction of RNF146 expression by chlorogenic acid (fig. S2E, F). Moreover, suppression of 6-OHDA-stimulated PARP activation afforded by chlorogenic acid was abolished when Akt1 activation was blocked by MK-2206 treatment (fig. S2E, F).

Fig. 3. Akt1-CREB activation by chlorogenic acid contributes to RNF146 expression.

(A and B) Western blotting analysis of Akt1 and CREB activation in DMSO- or chlorogenic acid-treated SH-SY5Y cells. Densitometry of phosphorylated protein was normalized to that of the protein’s total fraction; densitometry of RNF146 was normalized to that of β-actin. n = 3 separate experiments per group. (C and D) Western blotting analysis of total CREB, phosphorylated CREB (pCREB), and RNF146 in SH-SY5Y cells transfected with HA-tagged Akt1-CA, Akt1-DN, or mock as control. (n = 3 separate experiments per group). (E and F) Western blotting analysis of RNF146 abundance in SH-SY5Y cells with CRISPR-Cas9-mediated deletion of endogenous CREB. β-actin served as a loading control. n = 3 separate experiments per group. (G) Schematic diagram showing the promoter structures of luciferase constructs: Wild-type RNF146 promoter (pGL3-RNF146-Luc, WT), deletion mutants lacking CRE1 (ΔCRE1), CRE2 (ΔCRE2), or both (ΔCRE1+2). CRE1 and CRE2 motif sequences are indicated. TSS, transcription start site. Luc, luciferase. (H) Quantification of relative RNF146 promoter activity in SH-SY5Y cells transfected with the indicated luciferase constructs [refer to (G)] and pRL-TK (for 61 hours) and treated with chlorogenic acid (10 μM for 37 hours) or DMSO. n = 6 separate experiments per group. (I) ChIP of putative CREB response elements (CRE motifs 1 and 2) within the RNF146 promoter region determined by PCR. The promoter’s non-CRE region (Ctrl motif) and the β-actin region were used as negative controls; histone antibody or rabbit IgG were assay controls. Data are mean ± SEM; *P < 0.05, **P < 0.01, and ***P < 0.001 by ANOVA test followed by Tukey’s post hoc analysis.

Akt1 activation has been reported as an endogenous cell survival mechanism (15). CREB mediated transcription plays important roles in reprogramming the survival gene network in response to ischemic preconditioning stimuli (27). In addition, RNF146 expression was first identified in a preconditioning paradigm where cells reprogrammed cellular survival gene networks upon brief exposure to sublethal doses of toxins (8, 10, 22). We hypothesized that Akt1 activation could be involved in preconditioning stimuli-induced RNF146 expression. Supporting this notion, brief exposure to low-dose hydrogen peroxide induced RNF146 expression and led to robust Akt1 and CREB phosphorylation (fig. S3A, B). Activation of the CREB-RNF146 pathway during preconditioning is mediated by Akt1 activation as evidenced by the fact that MK-2206 treatment blocked preconditioning activation (fig. S3A, B). These results suggest that Akt1 activation during preconditioning contributes to RNF146 expression, thus rendering reprogrammed cells resistant to subsequent cytotoxic stimuli.

Next, we attempted to determine whether CREB is critical for RNF146 expression downstream of chlorogenic acid treatment. CREB knockdown was performed by transfecting SH-SY5Y cells with the CRISPR-Cas9 system expressing both gRNA targeting CREB and cas9. CRISPR-Cas9 successfully suppressed CREB expression, which ultimately blocked chlorogenic acid-induced RNF146 expression (Fig. 3E, F). It can be stated that CREB signaling is required for RNF146 regulation, downstream of chlorogenic acid treatment. To analyze RNF146 promoter motifs required for chlorogenic acid-mediated activation, we cloned RNF146 promoter luciferase constructs with deletions in CRE1 and/or CRE2 motifs (Fig. 3G). Chlorogenic acid-induced RNF146 promoter activation was substantially reduced by deletion of the CRE1 motif, while CRE2 deletion decreased RNF146 promoter activity to a lesser extent (Fig. 3H). Deletion of both CRE motifs showed repression of RNF146 promoter activity comparable to CRE1 deletion alone (Fig. 3H). Direct binding of RNF146 promoter by CREB was also monitored by anti-pCREB chromatin immunoprecipitation (ChIP), followed by PCR for CRE motifs on RNF146 promoter. ChIP revealed that phospho-CREB binding was enriched within regions of putative CRE motifs predicted by JASPAR. Binding of pCREB to CRE1 and CRE2 was increased by chlorogenic acid treatment, with CRE1 showing more enrichment than CRE2 (Fig. 3I). No pCREB binding was observed when using either the non-CRE motif within the RNF146 promoter or the β-actin region as the control (Fig. 3I).

Akt1-CREB pathway activation is required for PARP suppression and cytoprotection by chlorogenic acid

Based on the role of Akt1-CREB pathway activation in neuroprotective E3 ligase RNF146 expression, the cell protective function of this pathway was further assessed in a cellular PD model of 6-OHDA toxicity. Expression of Akt1-DN abolished cytoprotection by chlorogenic acid against 6-OHDA toxicity (Fig. 4A), indicating the requirement of Akt1 activation for chlorogenic acid-mediated cell survival. Chlorogenic acid treatment resulted in a similar level of cytoprotection against 6-OHDA-induced cell death compared to Akt1-CA expression alone (Fig. 4A). Supporting the role of PAR regulation in mediating parthanatic cell death, we observed correlative alterations in PAR levels with cytotoxicity profiles (Fig. 4A-C). Substantial PAR elevation induced by 6-OHDA was blocked by chlorogenic acid, an effect that was abolished by expression of Akt1-DN (Fig. 4B, C). Akt1-DN expression also repressed RNF146 induction by chlorogenic acid treatment (Fig. 4B, C). In contrast, Akt1-CA overexpression was sufficient to increase RNF146 expression and repress the 6-OHDA–induced increase in PAR abundance, even in the absence of chlorogenic acid treatment (Fig. 4B, C). These results indicate that Akt1 activation by chlorogenic acid is important in the induction of RNF146 expression and cytoprotection against PARP-dependent cell toxicity.

Fig. 4. Chlorogenic acid-mediated cytoprotection requires Akt1 and CREB activation.

(A) Trypan blue exclusion viability assay, in SH-SY5Y cells expressing Akt1-DN or Akt1-CA, monitoring the cytoprotective effect of chlorogenic acid (10 μM pre-treatment for 37 hours) against 6-OHDA-induced (70 μM for 16 hours) cell death. n = 6 separate experiments per group. (B and C) Immunoblotting analysis of levels of PAR-modified proteins, RNF146, and HA-tagged Akt1 in SH-SY5Y cells transfected and treated as indicated. 6-OHDA: 70 μM for 30 min; chlorogenic acid pre-treatment: 10 μM for 24 hours. Densitometry was normalized to that of β-actin. n = 3 separate experiments per group. (D) Trypan blue exclusion viability assay in SH-SY5Y cells transfected with CREB-targeted or control CRISPR-Cas9 gRNA and treated with chlorogenic acid (10 μM pre-treatment for 37 hours) then 6-OHDA (70 μM for 16 hours). n = 6 separate experiments per group. (E to G) Immunoblot analysis of levels of PAR-modified proteins, RNF146, and CREB protein in total lysates from cells described in (D). Densitometry was normalized to that of β-actin. n = 3 separate experiments per group. Data are mean ± SEM; *P < 0.05, **P < 0.01, and ***P < 0.001 by ANOVA test followed by Tukey’s post hoc analysis.

The role of CREB downstream of Akt1 was also determined with regard to cell viability and PAR regulation. When CREB expression was knocked down by the CRISPR-Cas9 system, chlorogenic acid failed to protect SH-SY5Y cells from 6-OHDA toxicity (Fig. 4D). Moreover, on 6-OHDA treatment in CREB-knockdown cells transfected with the gRNA-CREB construct, chlorogenic acid failed to increase RNF146 expression and repress the increase in PARylation (Fig. 4E-G). These results suggest that CREB is involved in mediating the effect of chlorogenic acid on RNF146 expression and suppression of PARP activity.

Chlorogenic acid activation of Akt1-RNF146 pathway represses parthanatos in 6-OHDA–treated dopaminergic neurons

We further monitored Akt1-RNF146 activation and cytoprotection by chlorogenic acid in dopaminergic neurons differentiated from neural progenitor cells (28). To label dopaminergic neurons, a reporter plasmid expressing GFP under the control of a mini TH promoter (29) was transfected into proliferating neural progenitor cells. Following dopaminergic neuron differentiation, GFP-labeled neurons exhibited increases in both RNF146 and pAkt1 expression in response to chlorogenic acid treatment (fig. S4A, B). Even under 6-OHDA-induced stress, chlorogenic acid led to comparable elevations of both pAkt1 and RNF146 expression (fig. S4A, B). Consistent with this activation of the Akt1-RNF146 pathway by chlorogenic acid, 6-OHDA-induced DNA fragmentation, monitored by TUNEL staining (30), was markedly reduced in dopaminergic neurons treated with chlorogenic acid (fig. S4C, D). Similarly, chlorogenic acid blocked substantial increases of PAR-modified protein in dopaminergic neurons under 6-OHDA-induced stress (fig. S4E, F). These data suggest that the Akt1-RNF146 signaling pathway in postmitotic neurons is stimulated by chlorogenic acid; hence, chlorogenic acid might be used for neuroprotection against oxidative stress.

RNF146 expression by chlorogenic acid prevents PARP activation and dopaminergic neuronal loss in a PD mouse model

RNF146 has been shown to be neuroprotective against neuronal death in diverse mouse models of stroke or PD. Thus, we sought to determine the potential application of chlorogenic acid in regulating RNF146 expression in the mouse brain and, thus, provide neuroprotection in PD mouse models. Intraperitoneal administration of chlorogenic acid increased RNF146 mRNA levels by two-fold in the mouse ventral midbrain (VM) (Fig. 5A). There was also an approximate three-fold increase in RNF146 protein expression in the VM of mice treated with chlorogenic acid (Fig. 5B, C). Moreover, consistent with the results from SH-SY5Y cells, RNF146 expression positively correlated with an increase in phosphorylation of both Akt1 and CREB in VM following chlorogenic acid administration (Fig. 5B, C). Dopaminergic expression of RNF146 by chlorogenic acid was assessed by co-immunolabeling of RNF146 and TH in VM. We observed an increase in RNF146 signal intensities in the TH-positive dopaminergic neurons of the VM of mice administered with chlorogenic acid (Fig. 5D, E).

Fig. 5. Akt1-CREB activation and dopaminergic RNF146 expression by chlorogenic acid administration in vivo.

(A) Relative Rnf146 mRNA expression in the VM of mice administered with chlorogenic acid (10 mg/kg i.p. for 3 days) or vehicle control. n = 6 mice per group. (B and C) Immunoblot analysis of Akt1, pAkt1, RNF146, CREB, and pCREB in the VM of mice described in (A). Densitometry of phosphorylated protein was normalized to that of the protein’s total fraction; that of RNF146 was normalized to β-actin. n = 6 mice per group. (D and E) Immunofluorescence imaging and quantitative analysis of RNF146 abundance in TH-positive dopaminergic neurons of mice treated as described in (A). Scale bar = 10 μm. n = 6 mice per group. Data are mean ± SEM; ***P < 0.001 by unpaired two-tailed student’s t test.

Because RNF146 expression in dopaminergic neurons can enhance neuronal survival, we employed a 6-OHDA PD mouse model to validate the therapeutic efficacy of chlorogenic acid. When we assessed the viability of TH-stained dopaminergic neurons in the SNpc using unbiased stereological counting, an intrastriatal injection of 6-OHDA caused more than 50% neuron loss, which was largely attenuated by chlorogenic acid administration (Fig. 6A, B). Chlorogenic acid alone was not toxic to dopaminergic neurons, since we did not observe any dopamine cell loss in the chlorogenic acid-treated and vehicle-treated mice group (Fig. 6A, B). Consistent with the robust degeneration of dopaminergic neurons by 6-OHDA, there was a strong increase in the amount of PAR in the VM of mice that had received a 6-OHDA intrastriatal injection (Fig. 6C, D). Chlorogenic acid increased RNF146, pAkt1, and pCREB expression in the VM of mice, regardless of 6-OHDA intrastriatal injection (Fig. 6C, E). In addition to inducing RNF146 expression, chlorogenic acid completely inhibited 6-OHDA-induced PARP activation in the VM. (Fig. 6C, D). Immunofluorescence analysis further validated the observed increase in dopaminergic expression of RNF146 and pAkt1 by chlorogenic acid in both 6-OHDA PD mice and control mice (Fig. 6F-I, and fig. S5). Together, these results demonstrate the therapeutic potential of chlorogenic acid application in preventing dopamine neurodegeneration in PD mouse models.

Fig. 6. Chlorogenic acid prevents dopamine cell loss and PARP activation in a PD mouse model.

(A) Representative TH immunohistochemical staining of the substantia nigra of mice pre-treated with chlorogenic acid (10 mg/kg i.p. for 7 days) or DMSO prior to stereotactic striatal injection of 6-OHDA (8 μg for 4 days; coordinates from bregma, L: −2.0, AP: 0.5, DV: −3.0 mm). Scale bar = 100 μm. (B) Stereological assessment of TH-positive dopaminergic neurons in the SNpc of the injection side from mice described in (A). n = 4 mice per group. (C to E) Western blotting analyses of the abundance of PAR, RNF146, phosphorylated Akt1 (pAkt1), total Akt1, pCREB, total CREB, and TH in the VM of 2-month-old mice pre-treated with chlorogenic acid (10 mg/kg/day i.p. for 4 days) or DMSO vehicle followed by intrastriatal 6-OHDA injection (for 37 hours). Densitometry of phosphorylated protein was normalized to that of the protein’s total fraction; that of PAR and RNF146 was normalized to β-actin. n = 4 mice per group. (F to I) Immunofluorescence analyses of the relative abundance of RNF146 (F and H) and pAkt1 (G and I) in TH-positive dopaminergic neurons of control (PBS-injected) and PD-model mice [as described in (A)]. Scale bar = 10 μm. n = 30 cells from 4 mice per group. Data are mean ± SEM; **P < 0.01 and ***P < 0.001 by unpaired two-tailed student’s t test or ANOVA test followed by Tukey’s post hoc analysis.

An additional PD mouse model with α-synucleinopathy was used to further examine dopaminergic neuroprotection by chlorogenic acid. We used a combinatorial nigral injection PD model of α-synuclein PFF and rAAV-αSyn (31) that displayed efficient viral transduction of entire VM dopaminergic neurons (Fig. 7A, and fig. S6A). α-Synuclein PFF was prepared by aggregation of mouse purified α-synuclein recombinant protein (fig. S7A) and subsequent sonication into small size fragments (fig. S7A, B) for efficient cellular uptake. Pathogenic capacity of sonicated α-synuclein was validated by observation of pSer¹²⁹-αSyn positive neuronal aggregates and enhanced neurotoxicity in response to PFF treatment to primary cultured cortical neurons (fig. S7C–E). Efficient knockdown of mouse RNF146 by rAAV-shRNF146 was validated in mouse cortical neuron culture by Western blotting using RNF146 antibodies (fig. S6B, C). Combinatorial injection of PFF and rAAV-αSyn resulted in elevated anxiety in mice, as shown by their increased tendency to explore the periphery of the open-field arena (Fig. 7B, C). Post-administration of chlorogenic acid attenuated this anxiety phenotype in combinatorial PD mice, while knockdown of endogenous mouse RNF146 by rAAV-shRNF146 coinjection abolished the protective effect of chlorogenic acid (Fig. 7B, C). Moreover, impairment of motor coordination in combinatorial PD mice, as observed in the rotarod test, was also reversed by chlorogenic acid administration (Fig. 7D). The protective effect of chlorogenic acid seems to be RNF146 dependent, since rAAV-shRNF146 coinjection blocked chlorogenic acid-induced improvement of motor impairment (Fig. 7D).

Fig. 7. Chlorogenic acid-mediated dopaminergic neuroprotection in a sporadic PD mouse model is dependent on the induction of RNF146 expression.

(A) Experimental scheme of combinatorial PD mouse model generation with coinjection of PFF (ventral tegmental area, VTA) and rAAV-α-synuclein (SNpc). rAAV-shRNA to mRNF146 was injected to ablate endogenous mouse RNF146 and rAAV-scrambled shRNA (rAAV-shControl) was used as an shRNA control injection. Experimental schedule is also presented as a diagram (bottom panel). IHC, immunohistochemistry. (B) Representative exploratory paths in an open-field test for mice with the indicated PFF/rAAV viral injections and drug treatments. (C) Assessment of anxiety in each experimental mouse group represented in (B), determined by the percentage of exploration time in the border versus the sum of the center and periphery zones. n = 4 mice per group. (D) Motor coordination of each experimental mouse group determined by the latency to fall in an accelerating rotarod test. n = 4 mice per group. (E) Representative TH immunohistochemical staining of substantia nigra from combinatorial PFF/rAAV-αSyn-injected PD mice treated subsequently with chlorogenic acid (10 mg/kg/d i.p. for 11 days) or DMSO. rAAV-shRNF146 was coinjected to ablate endogenous mouse RNF146 expression. Scale bar = 500 μm. (F) Stereological assessment of TH-positive dopaminergic neurons in the SNpc of the injection side from the indicated mouse groups. n = 4 mice per group. (G and H) Representative immunofluorescence images (G) and quantitative analysis (H) of RNF146 expression in TH-positive dopaminergic neurons of the indicated experimental mouse groups. Scale bar = 10 μm. n = 20 cells from 4 mice per group. Data are mean ± SEM; *P < 0.05, **P < 0.01, and ***P < 0.001 by unpaired two-tailed student’s t test or ANOVA test followed by Tukey’s post hoc analysis.

Next, dopaminergic neuron survival in combinatorial PD mice was monitored by TH stereological counting. Consistent with motor impairment, there were a 57% dopaminergic neuron loss (Fig. 7E, F), marked production of dopaminergic α-synuclein aggregates (fig. S8A, B), and neuroinflammation as evidenced by increased astrogliosis (fig. S8C, D) in combinatorial PD mice. Chlorogenic acid administration largely enhanced neuron survival in this PD mouse model in an RNF146-dependent manner (Fig. 7E, F). We also examined RNF146 expression in dopaminergic neurons of the SNpc from each experimental mouse group. Immunofluorescence analysis revealed more than four-fold increases of RNF146 expression in TH-stained dopaminergic neurons from mice treated with chlorogenic acid (Fig. 7G, H, and fig. S9). Even in conjunction with expression of PFF and rAAV-αSyn, there was an enhancement of RNF146 expression by chlorogenic acid, comparable to that observed in control mice (Fig. 7G, H, and fig. S9). In contrast, rAAV-shRNF146 coinjection efficiently knocked down RNF146 expression, even with chlorogenic acid treatment (Fig. 7G, H, and fig. S9).

The Akt-CREB-RNF146 pathway is dysregulated in PD pathogenesis

A dysfunctional Akt pathway has been implicated in the pathogenic death of dopaminergic neurons in PD (32). To extend this view to our findings from mouse and cell models, we investigated the potential dysregulation of Akt-CREB-RNF146 pathways in clinical PD samples. We assessed the protein abundance of RNF146 and the abundance and activation (phosphorylation) of Akt1 and CREB in postmortem striatal tissues from patients with PD and age-matched control subjects (Table 1) by Western blotting. These postmortem PD brain samples have an increased abundance of PARsylated proteins, indicating an overactivation of PARP1 (8). The observed loss of TH abundance in the striatum of the brains of patients with PD confirmed robust loss of dopaminergic neurons in the substantia nigra and dopamine axon terminal in the striatum (Fig. 8A). In this degenerating environment, levels of both total and phosphorylated Akt1 were reduced in patients with PD compared to those in age-matched control subjects (Fig. 8A, B). This result is consistent with a previous study that demonstrated dysfunctional Akt activation in the brains of patients with PD (32). We also observed reduction of both total and phosphorylated CREB protein levels in PD brains (Fig. 8A, B). Moreover, consistent with a previous report (8), RNF146 expression was nearly absent in striatal tissues of patients with PD, in contrast to relatively abundant RNF146 expression in striatal tissues of age-matched control subjects (Fig. 8A, B). We also examined Akt1-RNF146 alteration in postmortem temporal lobe brain tissues from patients with PD and pathologically confirmed Lewy bodies (Table 2). Immunofluorescence analysis revealed marked downregulation of Akt1 phosphorylation and RNF146 expression in the postmortem brains of patients with PD compared to that in age-matched control subjects (Fig. 8C, D). There was correlative expression of Akt1 phosphorylation and RNF146 expression among the postmortem brain tissues from patients with PD and age-matched control subjects (Fig. 8E), suggesting potential RNF146 regulation by Akt1 and the clinical relevance of this pathway in PD. These results from the postmortem brains of patients with PD suggest that the potential dysregulation of the Akt1-CREB pathway in PD may ultimately lead to insufficient expression of the neuroprotective E3 ligase, RNF146.

Table 1. Information on postmortem human striatum tissues used in this study.

For each of the postmortem striatal samples analyzed in this study, the donor’s sex, age, and pathology details are provided in the table. The postmortem interval (PMI, hours) is also noted. BRC#, Brain Resource Center number. N/A, information not available. AD, Alzheimer’s disease. PART, primary age-related tauopathy.

BRC#	Male/Female	Age	PMI	Pathology
2209	M	71	16	Control
2581	F	56	7	Control
2590	M	77	10	Control
2661	M	73	5.5	PD w/ dementia
2670	F	60	11	PD w/ dementia, AD possible
2680	F	81	11	PD w/ dementia, AD possible
2692	F	75	N/A	PD, PART

Fig. 8. Clinical relevance of Akt1-CREB-RNF146 signaling pathway in PD pathogenesis.

(A and B) Western blotting analysis of pAkt1, Akt1, pCREB, CREB, RNF146, and TH in postmortem striatum (STR) brain tissue samples from patients with PD and from age-matched control subjects (Ctrl). Densitometry was normalized to that of β-actin in each sample. n = 3 individual tissue samples per control and 4 per PD group. (C to E) Immunofluorescence analysis (C and D) and Pearson correlation plots (E) of the expression of RNF146 and pAkt1 in postmortem cortical sections from PD patients with cortical Lewy body pathologies and from age-matched control subjects (Ctrl). Scale bar = 10 μm. n = 24 cells from 6 controls, and 24 cells from 6 PD patients. Data are mean ± SEM; *P < 0.05, **P < 0.01, and ***P < 0.001 by unpaired two-tailed student’s t test or ANOVA test followed by Tukey’s post hoc analysis.

Table 2. Information on postmortem human temporal lobe samples used in this study.

The table lists the details known on each donor’s sex, age, race/ethnicity (all white American), and pathology for each of the postmortem-fixed temporal lobe samples analyzed in this study. The postmortem interval (PMI, hours) is also noted.

Donor ID#	Male/Female	Age	Race	PMI	Pathology
421	M	71	W	2	Control
557	M	80	W	3.5	Control
510	F	72	W	3	Control
818	F	68	W	4	Control
869	F	79	W	4	PD with dementia, Lewy-body-positive
1319	F	76	W	4.58	PD with dementia, Lewy-body-positive
3750	M	86	W	2.5	PD with dementia, Lewy-body-positive
942	F	89	W	3.5	PD with dementia, Lewy-body-positive

Discussion

In this study, we discovered a previously unknown mechanism of RNF146 expression that supports cell viability in DA neurons but is repressed in PD. Previously, RNF146 expression had been observed in response to preconditioning stimuli where cells developed endogenous protective genetic programs after brief exposure to sublethal doses of toxins or hypoxia. The only previously identified molecular mechanism of RNF146 expression was ER activation alongside ER binding to the RNF146 promoter (8). However, it was not clear whether preconditioning stimuli-regulated RNF146 expression was modulated by ER signaling or other unknown molecular pathways. We focused here on chlorogenic acid—an RNF146-inducing compound—and found that Akt1 phosphorylation of CREB promotes RNF146 expression and mediates cytoprotection against parthanatic (PARP1-dependent) cell death. Brief exposure of SH-SY5Y cells to low-dose oxidative stress induced CREB phosphorylation and RNF146 expression, which are dependent on Akt1 activation. Because the Akt1-CREB pathway has been implicated in other preconditioning paradigms (33), it is possible that RNF146 induction through Akt1 activation may contribute to enhanced cell survival capacity during preconditioning in more diverse organ systems and when confronted with toxic stimuli.

Direct activation of the Akt1-CREB pathway by chlorogenic acid increased RNF146 expression and suppressed 6-OHDA-induced PARP1 activation and cell death. Chlorogenic acid has been reported to possess reactive oxygen species (ROS) scavenging activity, which could have contributed to its amelioration of DNA damage-induced PARP1 activation. RNF146 induction seems to be critical in mediating cytoprotection against 6-OHDA-induced oxidative stress. This notion is also supported by the fact that the expression of Akt1-CA is sufficient to induce RNF146 expression and provide cytoprotection through a reduction in PARP1 activity. Moreover, both chlorogenic acid-mediated cellular protection and PARP1 inhibition were largely abolished by Akt1-DN expression, indicating that Akt1 activation by chlorogenic acid was the main inhibitor of parthanatos. Although we have shown that Akt2 activity was not altered after chlorogenic acid treatment in SH-SY5Y cells, further investigation is still needed to determine whether other Akt isoforms also mediate RNF146 induction and dopaminergic neuron survival by chlorogenic acid in vivo. We have also shown that WWE domain of RNF146 even without E3 ligase motif was sufficient to provide cytoprotection and PARP1 inhibition. Although this is unexpected result, it is consistent with cytoprotection and inhibition of AIF nuclear translocation afforded by RNF146-ΔRF. Further study would be required to determine molecular mechanisms of PAPR1 inhibition by RNF146-ΔRF.

One of the most significant aspects of our study is that we revealed the connection between Akt1 activation and regulation of the parthanatic cell death pathway. Although the role of Akt1 as an anti-apoptotic kinase has been extensively studied, its involvement in parthanatic cell death had not been investigated. RNF146 expression through Akt1 activation and CREB phosphorylation has been shown to be actively engaged in suppressing PARP1 overactivation, thus inhibiting parthanatos. This mechanism is analogous to transcriptional regulation of apoptosis by activated Akt1, which leads to FOXO1- and CREB-dependent expression of anti-apoptotic genes, which ultimately suppress apoptosis. Similarly, CREB-dependent regulation of RNF146 expression may be a significant anti-parthanatic event that can transcriptionally modulate cell death execution. It may be beneficial to screen for additional genetic alterations that might be involved in the regulation of the PARP1-dependent cell death pathway.

The therapeutic potential of chlorogenic acid has been studied in several neurodegenerative diseases including stroke, Alzheimer’s disease, and PD. An epidemiological study has been conducted on its neuroprotective function by intake of coffee, in which chlorogenic acid is abundant (34). The ROS scavenging activity of chlorogenic acid, direct binding of chlorogenic acid to Akt1, and activation of Akt1 through chlorogenic acid are beneficial mechanisms for PD treatment. Significant pathological alterations in PD include accumulation of oxidative stress and downregulation of Akt1 activity. In our study, dysregulation of the Akt1-CREB-RNF146 pathway was further confirmed using postmortem brains of patients with PD. Concomitant downregulation of RNF146 correlated with a substantial reduction of TH which is indicative of a loss of dopaminergic midbrain neurons. It is interesting to note that the Akt1-RNF146 pathway was repressed even in the postmortem cortex from PD patients. To ascertain clinical relevance of RNF146 dysregulation in PD pathogenesis and α-synucleinopathy, more extensive correlative analysis in a larger cohort of postmortem PD brains is necessary. Since chlorogenic acid can reactivate Akt1, which is repressed in PD, its application in cases of PD with ongoing neurodegeneration could help halt disease progression and protect dopaminergic neurons through increased RNF146 expression. It is important to note that Akt1 induction by chlorogenic acid has been shown to counteract not only apoptotic cell death but also parthanatic cell loss. The neuroprotection afforded by Akt1 activation may be more significant than other pharmacological strategies that target only one individual cell death pathway. Moreover, it is important to determine whether other Akt1-activating compounds that are known to be neuroprotective can also modulate RNF146 expression and parthanatos (35).

Materials and Methods

Chemicals and antibodies

Chlorogenic acid (cat# C3878), H₂O₂, 6-OHDA, and PARP inhibitor 3AB, were purchased from Sigma (Sigma-Aldrich, Merck KGaA). Tamoxifen (Selleck Chemicals) and Akt1 inhibitor MK2206 (#11593, Cayman Chemical) were also used. The following primary antibodies were used: mouse antibody to RNF146 (N201/35, 1:5000, NeuroMab), rabbit antibody to GFP (cat# 2956, 1:5000, Cell Signaling Technology), mouse antibody to PAR (for WB: cat# 4335-MC-100, 1:3000, Trevigen; for IF: cat# sc-56198, 1:3000, Santa Cruz Biotechnology), rabbit antibody to tyrosine hydroxylase (NB300-109, 1:2000, Novus Biologicals), rabbit antibody to CREB (cat# 9197S, 1:3000, Cell Signaling Technology), rabbit antibody to phosphorylated CREB (cat# 9198S, 1:3000, Cell Signaling Technology), mouse antibody to Akt1 (cat# 2967S, Cell Signaling Technology), rabbit antibody to pSer⁴⁷³-Akt1 (cat# 9018S, Cell Signaling Technology), rabbit antibody to Akt2 (cat# 5239, Cell Signaling Technology), rabbit antibody to pSer⁴⁷⁴-Akt2 (cat# 8599, Cell Signaling Technology), rabbit antibody to AIF (ab32516, Abcam), and rabbit antibody to tubulin β-3 (Tuj-1, #802001, 1:1000, BioLegend). The following secondary antibodies were used: peroxidase (HRP)-conjugated sheep antibody to mouse IgG (cat# RPN4301, 1:5000, GE Healthcare), HRP-conjugated donkey antibody to rabbit IgG (cat# RPN4101, 1:5000, GE Healthcare), biotin-conjugated goat antibody to rabbit IgG (cat# BA-1000, 1:1000, Vector Laboratories), HRP-conjugated mouse antibody to β-actin (cat# A3854, 1:10000, Sigma-Aldrich), Alexa fluor 568-conjugated donkey antibody to rabbit IgG (A10042, Invitrogen), Alexa fluor 488-conjugated goat antibody to rabbit IgG (A11008, Invitrogen), Alexa fluor 568-conjugated donkey antibody to mouse IgG (A10037, Invitrogen), Alexa fluor 488-conjugated donkey antibody to mouse IgG (A21202, Invitrogen), and Alexa Fluor 405-conjugated goat antibody to rabbit IgG (ab175652, Abcam).

Cell culture and treatments

Human neuroblastoma SH-SY5Y cells (ATCC) were grown in Dulbecco’s modified Eagle’s medium [high glucose (4.5 g/L) DMEM] containing 10% fetal bovine serum (FBS; vol/vol) and penicillin-streptomycin antibiotic solution (100 U/ml, Thermo Fisher Scientific). SH-SY5Y cells were grown at 37°C in a humidified atmosphere consisting of 5% CO₂ and 95% air. X-tremeGENE HP transfection reagents (Roche) were used for transient transfection following the manufacturer’s instructions (approximate transfection efficiency was 50 %). For the mouse dopaminergic neuron culture, neural precursor cells (NPCs) were isolated from ventral midbrains (VM) dissected from mouse embryos (E10~12), as previously described (28). Isolated NPCs were plated onto glass coated with poly-L-ornithine (30 μg/ml in DW overnight) and fibronectin (2 μg/ml in DW overnight) and then grown for expansion in N2 medium supplemented with b27 (Gibco), insulin (5 μg/ml, Sigma), basic fibroblast growth factor (bFGF; 20 ng/ml, R&D Systems), and epithelial growth factor (EGF; 20 ng/ml, R&D systems). During the 3 days of NPC expansion, old media was replaced by fresh media every day. Expanded NPCs were transferred to newly coated plates after cell dissociation by gentle pipetting in HBSS buffer. NPCs were then transfected using lipofectamine 3000 reagent (Thermo Fisher Scientific) and differentiated into dopaminergic neurons by culturing in N2 medium supplemented with b27, insulin, and ascorbic acid (0.2 mM, Sigma) for 5 to 7 days. For mouse primary neuron culture, CD1 mice were obtained from KOATECH Co. (Pyeongtaek, South Korea). Primary cortical neurons were prepared from E14 pups and cultured in Neurobasal media (Gibco) supplemented with B-27 (Gibco), 0.5 mM L-glutamine, penicillin and streptomycin (Invitrogen) on tissue culture plates coated with poly-L-lysine (Gibco). The neurons were maintained by changing medium every 2–3 days. α-Syn preformed fibril (PFF) was added at 7 days in vitro (DIV) and further incubated for 14 days for the TUNEL cell death assays or 7 days for the pSer¹²⁹-αSyn immunofluorescence experiments.

Plasmids

The RNF146 promoter luciferase reporter construct (pGL3-RNF146-Luc) (8), pLKO-shRNA-targeting RNF146, pLKO-shRNA-targeting dsRed, pEGFP-C2-human RNF146, and pEGFP-C2 have been previously described (10). We purchased plasmids expressing the constitutively active form of Akt1 (#10841, Addgene) and dominant negative form of Akt1 (#16243, Addgene). The plasmid pAAV2.5-THP-GFP (#80336, 2.5 kb rat TH promoter with downstream reporter GFP) was also obtained from Addgene to label differentiated dopaminergic neurons from neural precursor cells. The CRISPR-Cas9 construct targeting human CREB has been described previously (36). High-titer rAAV1-GFP-U6-m-RNF146 of in vivo injection quality with RNF146 knockdown validation and control rAAV1-expressing scrambled shRNA (rAAV1-GFP-U6-shRNA) were purchased from Vector Biolabs.

pGL3-RNF146-Luc reporters with deletions of either CRE1 and CRE2 motifs or both were generated by mutagenesis (QuikChange II XL Site-Directed Mutagenesis Kit, #200522, Agilent) with the following primers (5′– 3′): ∆CRE1, F- AATATGAAATATTGGGAATGGAAAAAATATGCTGACAAGTAGAGGGAC and R- GTCCCTCTACTTGTCAGCATATTTTTTCCATTCCCAATATTTCATATT; ∆CRE2, F- TGAGTGAGAAGGGGTTGAACTACAGTCTGTGCCTG and R- CAGGCACAGACTGTAGTTCAACCCCTTCTCACTCA. The integrity of the mutant plasmids was verified by sequencing.

Subcellular fractionation

SH-SY5Y cells were briefly washed with ice-cold PBS, and subjected for homogenization with dounce homogenizer (#357438, Weaton, 40 strokes) in the lysis buffer [300 mM mannitol, 0.1% BSA, 200 nM EDTA, and 10 mM HEPES in phosphate-buffered saline (PBS), pH 7.4] supplemented with protease (#786-108, G-biosciences)/phosphatase inhibitors (#78420, ThermoFisher Scientific). After incubation on ice for 10 min with vortexing every 3 min, the precleared lysates (50 × g for 10 min to eliminate unbroken cells) were centrifuged at 1,100 × g for 15 min. The supernatant and pellet were used to prepare post-nuclear, and nuclear fractions, respectively. The effective isolation of each subcellular fraction was confirmed with Western blots using antibodies to nuclear and cytosolic marker proteins (GAPDH for cytoplasm containing mitochondria, and PARP1 for nucleus).

Western blotting

SH-SY5Y cells were briefly washed with ice-cold phosphate-buffered saline (PBS) and total protein lysates were prepared in a lysis buffer (1% Nonidet P40 in PBS, pH 7.4) supplemented with protease/phosphatase inhibitors. After three freeze and thaw cycles in dry ice, total protein lysates were centrifuged at 14,000 × g for 30 min. The supernatants were mixed with 2X Laemmli buffer (Bio-Rad) supplemented with β-mercaptoethanol (Sigma-Aldrich) and boiled for 5 min. Proteins were separated by SDS-PAGE and transferred onto nitrocellulose membranes for immunoblotting. Immunoblotting was performed with the indicated antibodies, and bands were visualized by chemiluminescence (Pierce, Thermo Fisher Scientific). Densitometric analysis of the bands was performed using ImageJ software (NIH, http://rsb.info.nih.gov/ij/).

Luciferase assay

SH-SY5Y cells were co-transfected with pGL3-RNF146-Luc WT or deletion mutants and pRL-TK (Promega, Madison, WI, USA). Cells were harvested at the indicated days and lysates were assayed for firefly luciferase activity using the Dual-Luciferase Reporter Assay System (Promega) with a microplate luminometer (Berthold Technologies, Bad Wildbad, Baden-Württemberg), according to the manufacturers’ instructions. Firefly luciferase levels were normalized to those of the Renilla control. Cells treated with 0.1% DMSO were used as the negative control. Luciferase values for each group were normalized to that of the DMSO control.

Real-time quantitative PCR

Total RNA was extracted with QIAzol Lysis Reagent (cat# 79306, QIAGEN) and then treated with DNase I to eliminate genomic DNA contamination. cDNA was synthesized from total RNA (1.5 μg) using a first-strand cDNA synthesis kit (iScript cDNA synthesis kit, Bio-Rad). The relative abundance of target mRNA expression was analyzed using real-time PCR (QuantStudio 6 Flex Real-Time PCR System, Applied Biosystems,) with SYBR Green PCR Master Mix (cat# 4309155, Applied Biosystems). The relative mRNA expression levels of target genes were calculated by the ΔΔCT method (37) using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as an internal loading control. The primer sequences used for real-time PCR were as follows: hGAPDH: F, 5′-AAACCCATCACCATCTTCCAG-3′; R, 5′-AGGGGCCATCCACAGTCTTCT-3′. hRNF146: F, 5′-ATTCCCGAGGATTTCCTTGACA-3′; R, 5′-GCTCATCGTACTGCCACCA-3′. mGAPDH: F, 5′-TGGCCTTCCGTGTTCCTAC-3′; R, 5′-GAGTTGCTGTTGAAGTCGCA-3′. mRNF146: F, 5′-AGTCCTGTTCCAATACTGCACC-3′; R, 5′-GAAGCACCCTTTACACACAGAT-3′.

Chromatin immunoprecipitation

Chromatin immunoprecipitation was performed according to the manufacturer’s instructions (Millipore). Briefly, SH-SY5Y cells (treated with DMSO or chlorogenic acid) were fixed with 1% formaldehyde for 10 min at 37°C. Glycerol-quenched samples were lysed in 1 ml of SDS buffer containing protease inhibitors. The lysates were incubated for 10 min on ice and sonicated to shear DNA. The samples were centrifuged at 10,000 × g at 4°C for 10 min. Pre-cleared supernatant samples were incubated with anti-phosphorylated CREB, anti-histone antibodies, or rabbit IgG (rIgG)-agarose beads and washed several times. DNA was recovered by phenol-chloroform-ethanol purification after reverse cross-linking. PCR was performed with template input DNA and ChIPed DNA using the following primers: Putative phosphorylated CREB binding motif within RNF146 promoter CREB motif 1 (−1815 to −1804: CATCATGACTTA; F, 5′-TCCTGTGTCTTGCGTCCTAA-3′; R, 5′-AACACTCTTCCCCATCATGC-3′), CREB motif 2 (−1240 to −1229: GCTGACATCTTA; F, 5′-GGTGGGTTTGAACAAGGAAA-3′; R, 5′-TGTGCCAGGTGTTGCTCTAA-3′); RNF146 promoter-control region (F, 5′-GCGCAAGCATCACTGAACTA-3′; R, 5′-TGTTGCATTTTGGGATTTCA-3′); β-actin region (F, 5′-AGAGCTACGAGCTGCCTGAC-3′; R, 5′-AGCACTGTGTTGGCGTACAG-3′). CREB binding motifs within the RNF146 promoter were analyzed using JASPAR (http://jaspar.genereg.net/analysis) with matrix IDs MA0018.1 and MA0018.3 at the relative profile score threshold of 80%.

Cell viability assay

SH-SY5Y cells were plated in 6-well plates at a density of 0.5 × 10⁶ cells per well. SH-SY5Y cells were transfected with the indicated constructs and grown in DMEM containing low FBS (2.5%) with or without chemicals at the indicated concentrations. Cells were then harvested by trypsinization, washed twice with PBS, and then resuspended in serum-free DMEM. Resuspended cells were mixed with an equal volume of 0.4% trypan blue (wt/vol) and incubated for 2 min at room temperature. Live and dead cells were analyzed using a Countess II Automated Cell Counter (Life Technologies). Cell viability was also analyzed by CCK-8 assay (#CK04-11, Dojindo Molecular Technologies) according to the manufacturer’s instructions. For assessment of parthanatic and apoptotic neuronal death, TUNEL assay (#ab661110, Abcam) was used to label DNA fragmentation, according to the manufacturer’s guidelines.

Animal experiments

All animal experiments were approved by the Ethical Committee of Sungkyunkwan University (approval #, SKKU # 2019-04-22-2) and were conducted in accordance with all applicable international guidelines. Male C57BL/6N mice (3 months old) were purchased from Orient (Suwon, Korea). Animals were maintained in a 12-hour dark/light cycle in air-controlled rooms and were provided ad libitum access to food and water. Chlorogenic acid was administered to mice intraperitoneally. Daily chlorogenic acid administration (i.p., 10 mg/kg body weight) began on day 0 and was continued for 7 days, followed by stereological assessment of dopamine neuron counts. Intrastriatal injection of 6-OHDA was performed on day 3. For the combinatorial PD mouse model, chlorogenic acid (i.p., 10 mg/kg/day) was administered 10 days after nigral PFF/rAAV-αSyn injection; daily drug administration continued for 11 days in conjunction with subsequent behavioral testing and neuropathological analysis.

Behavioral tests

The open-field test consisted of a rectangular wood box (40 × 40 × 40 cm) divided into 64 (8 × 8) identical regions (5 × 5 cm). The field was subdivided into border, peripheral, and central sectors. The central sector included 4 central squares (2 × 2). The peripheral sector included the 12 squares surrounding the central sector. The remaining squares composed the border sector. The mouse was placed at the center of the open field and was allowed to explore it for 15 min under a dim light. The apparatus was cleaned with diluted 70% ethanol between each trial. A video tracking system (Smart v3.0 software, Panlab Harvard apparatus) was used to record the distance traveled as a measure of locomotor activity. The time spent in the center and the number of entries into the center were measured as anxiolytic indicators.

Prior to the rotarod test, the mice were trained for 2 days. On day 3, the mice were placed on an accelerating rotarod cylinder and the latency time at which they fell off the cylinder was measured. The speed of the rotarod was slowly increased from 4 to 40 rpm over the course of 5 min. A trial ended if the mouse fell off the rotarod. Test data are presented as means of the latency times from three trials.

Stereotaxic injection of 6-OHDA or PFF/rAAV-αSyn

For stereotaxic injection of 6-OHDA (8 μg), 3-month-old C57BL/6N mice were anesthetized with pentobarbital (60 mg/kg). The 6-OHDA injection procedure was performed as described previously (38), but with modifications. Briefly, an injection cannula (26.5 gauge) was inserted stereotaxically into the right-side striatum (anteroposterior, 0.5 mm from bregma; mediolateral, 2.0 mm; dorsoventral, 3.0 mm). 6-OHDA was infused at a rate of 0.2 μl/min, with a total of 2 μl of 6-OHDA (4 μg/μl in sterile PBS) injected into each mouse. The injection cannula was maintained in the striatum for an additional 5 min for complete absorption after the final injection of 6-OHDA. Scalp skin was sutured closed after the cannula was slowly removed from the mouse brain. Wound healing and recovery were monitored daily following surgery. Chlorogenic acid administration (i.p., 10 mg/kg/d) was started 3 days before 6-OHDA injection and continued for an additional 4 days. For stereological analysis, animals were perfused at 4 days after intrastriatal 6-OHDA injection and fixed intracardially with ice-cold PBS followed by 4% paraformaldehyde. Mouse brains were removed, cryoprotected in 30% sucrose, and processed for microtome sectioning and immunohistochemistry. The combinatorial α-synucleinopathy PD model was generated by injecting PFF (10 μg, VTA: AP −3.4, ML −0.5, DV −4.3) and rAAV-αSyn (human α-synuclein) (1 μl of titer 5 x 10¹¹ GC/ml, SNpc: AP −3.4, ML −1.3, DV −4.3). Similar surgical procedures to those used in the 6-OHDA PD model were followed. AAV-GFP, AAV-shRNF146, or scrambled AAV-shRNA (shControl) were also used for each experimental group. A pre-formed fibril (PFF) was prepared from pure recombinant mouse α-synuclein (Proteos, Inc, Kalamazoo, MI, USA) according to a previous report (39). Briefly, α-synuclein PFF was prepared in PBS by constantly agitating α-synuclein with a thermomixer (1,000 rpm at 37° C) (ATTO WSC-2630, Tokyo, Japan). After 7 days of incubation, the α-synuclein aggregates were diluted to 0.1 mg/ml with PBS and sonicated for 2 min (1 sec pulse on /off, keep on ice for 1 minute after 15 times) at 20 % amplitude (Sonics & Materials INC. VCX130, Newtown, CT). Before use, sonicated PFF was evaluated by western blotting, transmission electron microscopy, and functional assays in cortical neuron culture (pS129-αSyn induction, and neurotoxicity). Chlorogenic acid administration (i.p., 10 mg/kg/d) began 10 days after PFF/rAAV injections and was continued for an additional 11 days.

Transmission electron microscopy of PFF

10 μL of the PFF (5 μg/μl sonicated and unsonicated PFF in PBS) samples were pipetted onto a 200 mesh copper grids (EMS) with carbon-coated formvar film and incubated for 2 min. Excess liquid was removed by blotting. The grid was briefly placed on 10 μL of 2% uranyl acetate (w/v; Merck, Darmstadt, Germany), followed by blotting to remove excess liquid. This last step was repeated. Grids were allowed to dry before imaging on a Phillips CM 120 TEM operating at 80 kV. Images were captured and digitized with an US1000 CCD (2048 x 2048 pixel, 14 μm pixel, 100% fill factor) and 2k x 2k CCD camera (Gatan, Inc. USA) by advanced microscopy techniques.

Preparation of tissues for immunoblotting

Mice were euthanized by cervical dislocation. Mouse brain subregions such as VM were identified according to procedures described previously (40). The dissected mouse brain tissues were homogenized in lysis buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.5% Nonidet P-40, 10 mM Na-β-glycerophosphate, Phosphate Inhibitor Cocktails I and II [Sigma-Aldrich], and a complete protease inhibitor mixture [Roche]) using a Diax 900 homogenizer (Heidolph, Schwabach, Germany). The homogenized brain samples were rotated at 4°C for 30 min for complete lysis and then centrifuged at 52,000 rpm for 20 min. The supernatants were collected, and the protein levels were quantified using the BCA Protein Assay Kit (Pierce) with bovine serum albumin (BSA) standards. Proteins were then subjected to immunoblotting with the indicated antibodies. Immunoreactive bands were visualized with an enhanced chemiluminescence kit (Pierce). Densitometric analysis of protein bands was performed using ImageJ software.

TH stereological cell counting

After intracardial fixation of mice, the brains were post-fixed overnight with 4% paraformaldehyde and subsequently cryoprotected overnight in 30% sucrose in PBS (wt/vol). Coronal sections (40 μm thickness) were cut through the brain including the substantia nigra. Every fourth section was used for analysis. For analysis of TH-positive dopaminergic neurons, sections were incubated with rabbit polyclonal anti-TH antibody (1:1000 dilution) (Novus), followed by sequential incubations with biotinylated goat anti-rabbit IgG and streptavidin-conjugated horseradish peroxidase (HRP) using a Vectastain ABC kit (Vector Laboratories, Burlingame, CA) according to the manufacturer’s protocol. TH-positive cells were visualized using 3,3-diaminobenzidine (DAB, cat# D4293, Sigma-Aldrich) as an HRP substrate. Immunostained brain sections were counterstained with Nissl stain. The total number of TH-positive neurons in the SNpc was determined using the Optical Fractionator probe from the Stereo Investigator software package (MicroBrightfield, Williston, VT, USA). All stereological counting was performed with the counter blinded to each mouse treatment.

Immunofluorescence

Cells fixed with 4% paraformaldehyde in PBS, fixed mouse brain samples, or postmortem human temporal lobe sections were blocked with a solution containing 5% normal donkey serum (Jackson Immuno Research Laboratories, West Grove, PA, USA), 2% BSA (Sigma-Aldrich), and 0.1% Triton X-100 (Sigma-Aldrich) for 1 h at room temperature. Samples were then incubated with combinations of primary antibodies against RNF146, pAkt1, TH, or PAR—depending on the experiment—at 4°C overnight. Then, the fixed samples were washed with PBS containing 0.1% Triton X-100 and incubated with corresponding secondary antibodies conjugated with fluorescent dye at room temperature for 1 h. Fluorescent images were obtained using a fluorescence microscope (Axiovert 200 M, Zeiss, Oberkochen, Germany) or confocal microscope (LSM 710, Zeiss).

Statistical analyses

Quantitative data are presented as mean ± SEM. Statistical significance was assessed using either unpaired two-tailed Student’s t-test (two-group comparisons) or ANOVA test with Tukey’s HSD post hoc analysis (comparisons of more than three groups). Differences with a P value of <0.05 were considered statistically significant. GraphPad Prism software was used for preparation of all plots and all statistical analyses.