A Proteome-Wide Effect of PHF8 Knockdown on Cortical Neurons Shows Downregulation of Parkinson’s Disease-Associated Protein Alpha-Synuclein and Its Interactors

Synaptic dysfunction may underlie the pathophysiology of Parkinson’s disease (PD), a presently incurable condition characterized by motor and cognitive symptoms. Here, we used quantitative proteomics to study the role of PHD Finger Protein 8 (PHF8), a histone demethylating enzyme found to be mutated in X-linked intellectual disability and identified as a genetic marker of PD, in regulating the expression of PD-related synaptic plasticity proteins. Amongst the list of proteins found to be affected by PHF8 knockdown were Parkinson’s-disease-associated SNCA (alpha synuclein) and PD-linked genes DNAJC6 (auxilin), SYNJ1 (synaptojanin 1), and the PD risk gene SH3GL2 (endophilin A1). Findings in this study show that depletion of PHF8 in cortical neurons affects the activity-induced expression of proteins involved in synaptic plasticity, synaptic structure, vesicular release and membrane trafficking, spanning the spectrum of pre-synaptic and post-synaptic transmission. Given that the depletion of even a single chromatin-modifying enzyme can affect synaptic protein expression in such a concerted manner, more in-depth studies will be needed to show whether such a mechanism can be exploited as a potential disease-modifying therapeutic drug target in PD.


Introduction
Parkinson's disease is characterized by not only motor but also cognitive symptoms [1], for which there is currently no cure. The molecular bases underlying these cognitive symptoms of PD are complex and may involve deficits in both pre-synaptic and postsynaptic transmission, which are thought to occur early in the course of the disease and therefore may serve as potential therapeutic targets [2].
Single point mutations in an epigenetic modifying enzyme can be responsible for devastating consequences, including the inability to form long-term memories [3]. Here we investigate the function of such an enzyme: PHF8 is a histone demethylase that has been implicated in two conditions: the first being X-linked intellectual disability (XLMR), a condition characterized by a profound loss of memory formation, and the second being Parkinson's disease [4]. Previously, we found evidence that PHF8 regulates the activityinduced expression of the neuronal protein ARC [5], a major regulator of synaptic function [6]. Synaptic dysfunction is indeed found in early stages of PD neuropathology and is mediated by apparent pathological elevation of Alpha-synuclein (SNCA), a PD-implicated protein that is central to the synaptic dystrophy that precedes dopaminergic cell loss [7], Table 1. A manually curated shortlist of the top-ranked proteins that are downregulated when PHF8 is knocked down in activated neurons showing a functional enrichment of synaptic plasticity and Parkinson's disease. The full set of hits can be found in Supplementary Table S1. The genes encoding for the proteins that are in bold are known to be potential target genes of PHF8 as per the ChIP-Atlas [19].

Primary Culture of Cortical Neurons
Cortices were dissected from E18 embryos of Sprague-Dawley rats (Rattus norvegicus), which were then subjected to the Papain Dissociation System (Worthington Biochemical Corporation, Lakewood, CA, USA). Dissociated cells were plated on poly-D-lysine-coated dishes at a plating density of 1.5 × 105/cm 2 in neurobasal medium (Gibco, Grand Island, New York, NY, USA) supplemented with 10% (v/v) fetal bovine serum (FBS), 1% (v/v) penicillin-streptomycin (P/S, Gibco, Grand Island, New York, NY, USA), and 2% (v/v) B27 supplement (Gibco, Grand Island, New York, NY, USA) for 2 h. Medium was changed on days in vitro (DIV) 5. Subsequently, medium was changed every three to four days. All experiments were carried out on DIV 21 as previously described [5].

Pharmacological Stimulation of Neural Network Activity
Primary rat cortical neuronal cultures were treated with a combination of 100 µM 4-aminopyridine (4AP), 50 µM bicuculline, and 50 µM forskolin (hereafter abbreviated as 4BF) for 8 h to induce neural network activity as previously described [4]. This protocol has been previously reported to induce pharmacological LTP [66,67].

Cell Lysate Preparation and RNA/Protein Extraction
Following 4BF stimulation, neuronal cultures were washed gently with phosphatebuffered saline (PBS). Cells were gently scraped off and harvested in an Eppendorf tube. Cells were spun down at 10,000× g for 5 min at 4 • C to obtain the cell pellet. Total protein was isolated using an RNA/protein extraction kit (Macherey-Nagel, Düren, North Rhine-Westphalia, Germany) as specified by the manufacturer. A BCA kit (Pierce, Rockford, IL, USA) was used to measure the concentration of proteins. A total of 30 µg of each protein sample was used for subsequent mass spectrometry experiments. For qRT-PCR studies, 4BF-treated neuronal cells were washed, scraped, and spun down as above. RNA samples were obtained from the cell pellet using the RNA-protein extraction kit as specified by the manufacturer (Macherey-Nagel, Düren, North Rhine-Westphalia). cDNA was synthesized and subsequently purified using a spin column (Qiagen), then eluted into 30 µL volumes; thereafter, 2 µL was used in qRT-PCR employing SYBR Green using primers against the transcriptional start site (TSS) of known activity-regulated neuronal genes such as Arc (NCBI Gene ID: 54323), BDNF (NCBI Gene ID: 24225), Fos (NCBI Gene ID: 314322), and control genes including Rpl19 (NCBI Gene ID: 81767) and GAPDH (NCBI Gene ID: 24383). The primers used for PHF8 qRT-PCR were CCTAAAGCCCGTGTGACT and GGCGCGGCTGTTCTACCT. Statistical analyses were performed using Student's two-tailed t-test with a p-value < 0.05 being considered significant.

Isobaric Tag for Relative and Absolute Quantification of Proteins (iTRAQ)
Following reduction and alkylation, proteins collected from primary cortical neurons were trypsinized overnight at 37 • C. Peptides were dried and resuspended in massspectrometry-compatible buffer and then labeled with the 8-plex iTRAQ labeling reagent (Applied Biosystems, Foster City, CA, USA). Labeled samples were combined and analyzed with one-dimensional nanoLC-MS/MS (Dionex UltiMate 3000 nanoLC system coupled with AB Sciex TripleTOF 5600 system) for protein identification. The IPI human protein database (version 3.77) was searched using ProteinPilot (version 4.5, AB Sciex, Framingham, MA, USA) and the identified hits were analyzed using DAVID (http: //david.abcc.ncifcrf.gov; accessed on 11 January 2023) for gene ontology annotation [68]. The mass spectrometry proteomics data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD036335 [69].

Immunofluorescence
Primary cortical neuronal cells transfected with expression vectors or shRNAs as indicated were pre-extracted with 100% methanol or directly fixed for 5 min with 4% formaldehyde in sucrose buffer, then permeabilized for 2 min with 0.1% Triton X-100 if there was no pre-extraction. After three rinses with PBS/0.1% Triton X-100, blocking solution (10% BSA and goat serum in PBS, pH 7.4) was applied for 30 min and primary antibodies were added in blocking buffer for 1 h at room temperature. After three 5 min washes with PBS/0.1% Triton X-100, cells were incubated with secondary antibodies conjugated with fluorescent dyes (Alexa Fluors, Invitrogen) for 1 h, washed again with PBS/0.1% Triton X-100, and mounted in 97% thiodiethanol in PBS (Sigma-Aldrich, St. Louis, MO, USA). Images were recorded on a Nikon-Ti microscope (Nikon, Tokyo, Japan) equipped with an Andor camera (Andor, Belfast, Ireland) at 1 × 1 binning and a 60× objective (Nikon). Z-stacks (0.2 µm sections) were deconvolved using AutoQuant (Nikon NIS Elements) and projected for maximum intensity. Image intensities for each antibody were scaled identically.

Widefield Imaging Microscopy
Fluorescence images were obtained using a motorized inverted wide-field epifluorescence microscope (Nikon Eclipse Ti-E) using the 20× objective lens. Motorized excitation and emission filter wheels (Ludl electronics, Hawthorne, NY, USA) fitted with a DAPI/CFP/YFP/DsRed quad filter set (#86010, Chroma, Rockingham, VT, USA) were used together with filter cubes for DAPI, CFP, YFP, and TxRed (Chroma) to select specific fluorescence signals. Z-stacks were obtained spanning the entire nucleus and out-of-focus fluorescence was removed using the AutoQuant deconvolution algorithm (Media Cybernetics). Images were digitized using a cooled EM-CCD camera (iXon EM+ 885, Andor, Belfast, Northern Ireland). Image acquisition was performed using NIS Elements AR 4.2 software (Nikon). NIS Elements Binary and ROI Analysis tools were used to segment nuclei based on DAPI signal intensity.

Overexpression of PHF8-YFP and Knockdown of PHF8 Levels Using shRNA and siRNA Transfection
For the short-hairpin RNA (shRNA) experiments, a combination of two shRNA plasmids targeting the PHF8 sequence were used: RLGH-EN02366 and RLGH-EU01979 (Transomic Technologies, Huntsville, AL, USA). Neuronal cultures (DIV21) were transfected with either the shRNA plasmids or a plasmid containing PHF8 fused to Yellow Fluorescent Protein (YFP) using the Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. For the overexpression experiments, DNA containing PHF8-YFP construct that was previously created [4] was added to Lipofectamine at a ratio of 1:1. The Lipofectamine:DNA complex was incubated at room temperature for 20 min before being added to the cells. The complex was added dropwise such that it was evenly distributed on the cell culture. Culture medium was added after 20 min and experiments were performed on DIV21.
For the small interfering RNA (siRNA) experiments, a combination of three unique 27 mer siRNA duplexes targeting the PHF8 sequence were used (Locus ID 317425, Origene, Rockville, MD, USA) and transfected into primary cortical neuronal cells at DIV19 using the LipofectamineRNAiMax reagent according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA). After transfection, neuronal medium was changed, and neurons were allowed to equilibrate prior to sample collection at DIV21 for mass spectrometry analyses.

Relative Quantitation of the Activity-Regulated Cortical Neuronal Proteome
We performed the quantitative proteomics method of isobaric tag for relative and absolute quantitation (iTRAQ), which has been validated for protein quantitation in neural tissues [70,71]. Using iTRAQ, we analyzed primary cultured cortical neurons from Rattus norvegicus. After combining 2 biological replicates, we identified 2678 unique proteins though multidimensional protein identification (MudPIT), out of which 2656 proteins were confirmed with more than one peptide (Figure 1, Supplementary Table S1).

RNA Interference Specifically Depleted PHF8 in Primary Rat Cortical Neurons
We used a specific siRNA directed against rat PHF8 and scrambled siRNA as a control to transfect primary neuronal cultures. In order to that validate PHF8 knockdown was successful at the transcript level, neurons were subjected to RT-PCR 3 weeks after at DIV 21, after an 8 h stimulation protocol with 100 µM 4-Aminopyridine, 50 µM Bicuculline, and 50 µM Forskolin. Transcript levels of control genes such as the ribosomal protein Rpl19, the housekeeping gene GAPDH, and another transcriptional regulator p300 served as the negative controls ( Figure 2). Biomedicines 2023, 11, x FOR PEER REVIEW 7 of 14

RNA Interference Specifically Depleted PHF8 in Primary Rat Cortical Neurons
We used a specific siRNA directed against rat PHF8 and scrambled siRNA as a control to transfect primary neuronal cultures. In order to that validate PHF8 knockdown was successful at the transcript level, neurons were subjected to RT-PCR 3 weeks after at DIV 21, after an 8 h stimulation protocol with 100 μM 4-Aminopyridine, 50 μM Bicuculline, and 50 μM Forskolin. Transcript levels of control genes such as the ribosomal protein Rpl19, the housekeeping gene GAPDH, and another transcriptional regulator p300 served as the negative controls ( Figure 2).

RNA Interference Specifically Depleted PHF8 in Primary Rat Cortical Neurons
We used a specific siRNA directed against rat PHF8 and scrambled siRNA as a trol to transfect primary neuronal cultures. In order to that validate PHF8 knockdown successful at the transcript level, neurons were subjected to RT-PCR 3 weeks after at 21, after an 8 h stimulation protocol with 100 μM 4-Aminopyridine, 50 μM Bicucu and 50 μM Forskolin. Transcript levels of control genes such as the ribosomal pr Rpl19, the housekeeping gene GAPDH, and another transcriptional regulator p300 se as the negative controls ( Figure 2).

Validation of Alpha-Synuclein as a Synaptic Protein Regulated by PHF8
As alpha-synuclein (SNCA), a critical protein in Parkinson's disease pathogenesis, turned out to be amongst the main proteins identified via proteomics to be downregulated by PHF8 depletion (Table 1). In order to validate this iTRAQ finding, we transfected neurons with PHF8 to overexpress it, then contrasted these neurons against neurons transfected with a specific shRNA against PHF8 (Figure 3).
(siRNA)-values of 2 −delta(delta CT) normalized to a ribosomal gene Rpl19 in cortical neurons, which are quiescent (control) compared to neurons subjected to 3 h of 4AP + Bicuculline + Forskolin (4BF) with and without PHF8 siRNA, showing effective knockdown of the expression of PHF8 at 3 weeks (DIV 21) on neuronal and glial markers. Data are represented as mean ± SEM. All differences that are significant with a p*-value of <0.05 (Student's two-tailed t-test) were marked with an asterisk.

Validation of Alpha-Synuclein as a Synaptic Protein Regulated by PHF8
As alpha-synuclein (SNCA), a critical protein in Parkinson's disease pathogenesis, turned out to be amongst the main proteins identified via proteomics to be downregulated by PHF8 depletion (Table 1). In order to validate this iTRAQ finding, we transfected neurons with PHF8 to overexpress it, then contrasted these neurons against neurons transfected with a specific shRNA against PHF8 (Figure 3).

Pathway Analysis Using DAVID Reveals Specific Downregulation of Proteins in PHF8 Knockdown That Are Involved in Synaptic Function
Out of the 2656 unique proteins quantified using MudPIT (Supplementary Table S1), we manually curated 41 proteins that were selected based on their physiological function and role in human diseases (Table 1). Amongst the proteins that were downregulated by PHF8 depletion in an activity-dependent manner, at least 33 were involved in synaptic function (Table 1). A search using the functional enrichment analysis tool DAVID led to the discovery of top terms including "synaptic vesicle endocytosis", "modulation of synaptic transmission", and "synapse organization" (Table 2) [68].  red) showing that the overexpression of PHF8 does not seem to affect SNCA levels (the level of red signal did not change compared to other neurons that are not transfected with green and due to the mixture of red and green the cell has turned slightly yellow). Scale bar = 50 um. (B) Transfection of a shRNA plasmid against PHF8 results in a qualitative reduction of SNCA (the green neuron does not turn yellow; not quantified).

Pathway Analysis Using DAVID Reveals Specific Downregulation of Proteins in PHF8 Knockdown That Are Involved in Synaptic Function
Out of the 2656 unique proteins quantified using MudPIT (Supplementary Table S1), we manually curated 41 proteins that were selected based on their physiological function and role in human diseases (Table 1). Amongst the proteins that were downregulated by PHF8 depletion in an activity-dependent manner, at least 33 were involved in synaptic function (Table 1). A search using the functional enrichment analysis tool DAVID led to the discovery of top terms including "synaptic vesicle endocytosis", "modulation of synaptic transmission", and "synapse organization" (Table 2) [68].

The Potential Role of PHF8 in Synaptic Plasticity and PD Pathogenesis
Synaptic plasticity in the form of long-term potentiation (LTP) and long-term depression (LTD) may be dysregulated in Parkinson's disease even before motor symptoms start to manifest. In this mass-spectrometry-based proteomics study using primary rat cortical neurons, 2605 unique proteins that were identified as differentially regulated by PHF8, out of which the top 41 proteins are listed in Table 1, had functions in synaptic transmission ( Table 2). As this was the first attempt to assess the role of PHF8 in an in vitro model of PD pathogenesis, we were not able to compare our results against published data in the literature. Nonetheless, the protein targets curated in the current dataset can be crossmatched with published PHF8 genetic targets obtained by various groups via ChipSeq (Supplementary Table S2 [72]).
We previously found that PHF8 is required for the transcription of Arc, the master regulator of synaptic plasticity [5]. Now, using mass spectrometry coupled with the isobaric tag for relative and absolute quantitation (iTRAQ) method, we quantified proteome-wide effects of PHF8 depletion in primary cortical neurons stimulated using a combination of three pharmacological agents to induce chemical LTP [66,67] and report that PHF8 may play a regulatory role in the activity-induced expression of neuronal proteins such as Alphasynuclein and Calmodulin Kinase II Alpha, which in turn play critical roles in the function of neuronal synapses and the pathophysiology of PD. Pathway analysis using DAVID allowed for the identification of groups of proteins affected by PHF8 with distinct functions: with respect to LTP, the Calmodulin Kinase II Alpha-Calmodulin-dependent Kinase 2 axis was downregulated by PHF8 depletion ( Table 2, Supplementary Table S3). Multiple synaptic transmission pathways that have been implicated in PD pathogenesis are seen to converge as NSF, AMPH, EIF4E, and CHN1 are all downregulated by PHF8 ( Table 2). Specific to clathrin-mediated endocytosis of synaptic vesicles, the pre-synaptic protein auxilin (DNAJC6), synaptojanin-1 (SYNJ1), and endophilin A1 (SH3GL2) [10] were also affected by PHF8 downregulation. These subgroups of PHF8-regulated synaptic proteins show that the knockdown of a single epigenetic enzyme can have a highly concerted effect on synaptic function.

Knockdown of PHF8 Reduces Levels of Alpha-Synuclein
This proteomics study revealed that levels of SNCA reduced significantly with PHF8 knockdown, which we attempted to validate using immunofluorescence microscopy ( Figure 3). The effective reduction of SNCA levels by PHF8 knockdown raises the possibility of potentially using this mechanism as a potential therapeutic target by reducing the function of PHF8, either via gene therapy [73] or through pharmacological means, since PHF8 is an enzyme.

Knockdown of PHF8 Reduces the Levels of Other Synaptic Plasticity-Related Proteins and SNCA Interactors
In addition to CamKIIa, amongst the other synaptic proteins that were differentially regulated by PHF8 in primary cortical neurons, we highlight CPLX1, which is involved in stimulus-dependent control of synaptic vesicle exocytosis and is a known interactor of SNCA [14]. Another interesting target from the current proteomics dataset is Rabphilin-3A (RPH3A), which is implicated in levodopa-induced dyskinesia [74] and is dysregulated in PD. Finally, CHN1, which was identified in a gene network in PD and AD [75], was also upregulated by PHF8 knockdown in our dataset. SNCA is one of many possible target genes of PHF8, according to the publicly available ChIP-Atlas dataset [19], supplementing the current proteomic-level observation (Supplementary Table S2) [72].
In summary, the concerted downregulation of not just Alpha-synuclein, but also many of its interactors, which have various roles in synaptopathy by PHF8 knockdown, deserves further study. Further validation of the role of a single epigenetic modifying enzyme such as PHF8 in a potential disease-modifying therapy through modulation of synaptic function should be considered.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/biomedicines11020486/s1, Supplementary Table S1 lists the proteins that were differentially expressed in the PHF8-rich vs. PHF8-depleted sample. Supplementary  Table S2 lists the possible target genes of PHF8 based on ChipSeq data [52]. Supplementary Table S3 lists the full results of DAVID Gene Ontology search to map the proteins that were differentially expressed in the PHF8-rich vs. PHF8-depleted sample according to their biological function [68].  Institutional Review Board Statement: The animal study protocol was vetted according to the guidelines of the Institutional Animal Care and Use Committee (IACUC). The approved IACUC protocol was 2011/SHS660.

Informed Consent Statement: Not applicable.
Data Availability Statement: The mass spectrometry proteomics data have been deposited to the Pro-teomeXchange Consortium via the PRIDE [69] partner repository with the dataset identifier PXD036335.