Glial cell-specific proteomic data from the substantia nigra of a rat 6-OHDA and fluorocitrate model of astrocyte death and microglial activation

Abstract

The dataset shows proteomic results (timsTOF Pro 2 (Bruker)) obtained using an originally developed method of adult rat brain isolation of astrocytes or microglia from the same sample. Mechano-enzymatic dissociation and FACS sorting retrieved pure, separate cellular fractions from the substantia nigra. Results come from an animal model of early Parkinson’s disease of selective nigrostriatal dopaminergic system neuron degeneration by 6-OHDA, combined with 7-day-long astrocyte dysfunction and death induced by fluorocitrate. Astrocyte and neuron death both induce microglial activation, but to varying degrees and through different mechanisms. Previous studies did not allow for assigning changes in common mechanisms (such as, for example, energy metabolism) to a specific cell type in tissue, while in vitro studies lack functional dimension. This research enables the identification of clear information on mechanisms within each cell type, originating from a multidimensional environment, while maintaining the functional and tissue-specific context. Comparison of astrocyte death-induced vs neuron death-induced microglia activation processes can be analysed using this dataset. Raw data are available via ProteomeXchange with identifiers PXD066353 and PXD067265.

Specifications Table

Subject	Biology
Specific subject area	Neurobiology of glia-neuron interaction
Type of data	Chromatograms, Tables, Image, Raw, Filtered, Processed, Analysed
Data collection	Brain dissociated astrocytes or microglia were FACS sorted (BD FACSAria Fusion). Protein extracts were analysed (timsTOF Pro 2, Bruker), performing DDA proteomics, positive-ion mode. Peptides separated (Ultimate 3000 RSLC nano system (Thermo Scientific), bioZen C18 250 × 0.075 mm column (Phenomenex), trap cartridge Acclaim PepMap C18 5 × 0.3 mm (Thermo Scientific)) 2–35 % acetonitrile + 0.1 % formic acid gradient. Mass spectrometer equipped with IMS for analysis of ions in a 1/K0 range of 0.6 to 1.6. Data acquired over an m/z range of 100–1700, processed for protein identification and quantification with DDA settings for PASEF scan.
Data source location	Laboratory of Proteomics and Mass Spectrometry, Maj Institute of Pharmacology, Polish Academy of Sciences, Krakow, 12 Smetna St., 31–343 Krakow, Poland
Data accessibility	The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium (http://www.proteomexchange.org/) via the PRIDE (https://www.ebi.ac.uk/pride/) partner repository in two datasets, for microglia and astrocyte data, respectively. Project Name: Microglia proteome from the rat substantia nigra. Study in an animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation. Direct URL to data: https://www.ebi.ac.uk/pride/archive/projects/PXD066353/privatereviewdataset Project accession: PXD066353 Project DOI: 10.6019/PXD066353 Project Name: Astrocyte proteome from the rat substantia nigra. Study in an animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation. Direct URL to data: https://www.ebi.ac.uk/pride/archive/projects/PXD067265/privatereviewdataset Project accession: PXD067265 Project DOI: 10.6019/PXD067265
Related research article

1. Value of the Data

• •This dataset contributes as one of the few cell-type-specific proteomic resources from an in vivo Parkinsonian model. It shows separately data on pure astrocytes or microglia isolated directly from the adult rat brain.
• •The animal model used here was originally developed and validated by us and well described in the previous publications [1,3,4]. It was shown that the death of astrocytes induced by fluorocitrate caused massive microglia activation and stressed neurons. Concomitant astrocyte loss and microglia activation did not allow us to distinguish whether functional disability was due to astrocyte loss or rather strong microglia activation. To separate each cell type's cellular response mechanisms - loss of astrocyte function, microglia activation, neuron death- we used FACS cell sorting and proteomics. This data allows us to describe each cell type mechanism separately, at the same time keeping the broader context of a well-studied animal model with a full behavioural description.
• •Neuronal death is irreversible, while astrocytes can renew their population, which makes it difficult to document and detect transient astrocytic dysfunction in the human brain, especially at the early stages of the progressive neurodegenerative disease, before diagnosis. Therefore, the use of an animal model of glia dysfunction and death was necessary to specifically address the role of glia in Parkinson’s disease.
• •Glia activation has been considered as a downstream response to neuronal damage and alpha-synuclein aggregation, although loss of protective function or gain of toxic role of astrocytes and microglia were found to be relevant mechanisms [2]. Lack of the exact answer required further dedicated research. Many previous studies focus on either neurons alone or microglia or astrocytes independently, while it is important to look at the broader tissue context, including various cell type interactions. This study aimed to fill this gap.
• •There is scarce data available coming from a single cell type from in vivo studies, especially in Parkinson’s disease. In this study, a new method of isolation of astrocytes or microglia from the rat brain was developed. In contrast to other studies, already differentiated cells were isolated from adult rats, not puppies.
• •Obtained information comes from a very small structure substantia nigra, essential in Parkinson’s disease. Its specific function and cellular composition, and 3D connections in the brain, are key to understanding pathomechanisms of this neurodegenerative disease.
• •The other advantage of this method is that both pure fractions of astrocytes or microglia were isolated from the same animals, giving direct mechanistic input in the described processes and an exact, broader context. Therefore, those two large datasets are of high scientific value.
• •Proteins with changed expression should become targets for further, cell-specific research in neurodegenerative disease.

2. Background

Nervous system inflammation and glia activation are among the hallmarks of most neurodegenerative diseases and Parkinson’s disease. Currently, there is no cure to stop the progression of neurodegeneration. Therefore, astrocytes and microglia are considered new, potentially druggable targets for disease-modifying strategies. Partial nigrostriatal neurodegeneration at early disease stages can be functionally compensated. Interaction between neurons, microglia, and astrocytes could be essential for this adaptation and neuronal survival in the long term, but the mechanisms responsible are unknown.

The aim was to check how astrocyte vs neuron degeneration differentially affected microglia activation proteome and describe molecular processes in each cell type separately.

In the studied rat model, fluorocitrate (FC) was infused into the substantia nigra (SN) [1,3,4]. By inhibition of aconitase FC causes an energy deficit in astrocytes, which become dysfunctional and die, inducing activation of remaining astrocytes and microglia. 6-OHDA given in a low dose causes selective degeneration of dopaminergic neurons and only a slight microglia response. As a result of a medium-sized lesion, temporal motor dysfunction is reversed as described in previous studies of the same model [1,3]. Briefly, motor dysfunction in this model, in animals treated with 6-OHDA into the MFB, is observed 4 days post-lesion, manifested by the decreased walking path length, locomotion time (total, supported, and free rearing), and increased resting times. Most of the behavioural deficits were diminished already after 6 days. This compensation was blocked due to FC infusion, proving that astrocytic support was essential for the process. Further description of energy metabolism changes selective to astrocytes or other cells was published in [4]. The previous research gives a broad context of the studied model. Presented here new dataset that complements the cited previous work. Understanding the role of individual cell types in the relation to diseased tissue cellular context is essential to understand disease pathomechanisms and identify better pharmacological targets.

3. Data Description

The project included 4 experimental groups. From each sample (Table 1), astrocytes and microglia cells were isolated simultaneously and processed for proteomic analysis. List of raw data files for microglia can be found in Table 2 and for astrocytes in Table 3.

Table 1.

List of analysed samples, including the number of cells isolated from rat substantia nigra tissue and the volume of protein extracts obtained. From each protein extract, a proportion of 5000 microglia cells or 19,000 astrocytes, or the whole sample, was analysed using the timsTOF Pro 2 mass spectrometer.

Sample	Group	Number of microglia cells eluted to the sample	Volume of sample [µl]	Number of astrocyte cells eluted to the sample	Volume of sample [µl]
SS1	sham+sham	1 061	14	108 371	1387
SS2	sham+sham	2 121	27	21 005	269
SS3	sham+sham	1 138	15	50 608	648
SS4	sham+sham	993	13	36 059	462
SF1	sham+fluorocitrate	16 821	215	12 029	154
SF2	sham+fluorocitrate	5 963	76	3 370	43
SF3	sham+fluorocitrate	5 136	66	4 736	61
SF4	sham+fluorocitrate	2 544	33	17 018	218
LS1	lesion+sham	4 956	63	5 281	68
LS2	lesion+sham	6 158	79	7 155	92
LS3	lesion+sham	6 013	77	8 191	105
LS4	lesion+sham	6 285	80	2 659	34
LF1	lesion+fluorocitrate	13 423	172	4 487	57
LF2	lesion+fluorocitrate	22 715	291	5 179	66
LF3	lesion+fluorocitrate	16 341	209	3 280	42
LF4	lesion+fluorocitrate	27 946	358	2 904	37

A total of 57 files in the Microglia dataset include raw data (*.mzML), peak analysis (*.mgf), and results (*.mzid). Samples SF3, SF5, SF7 are additional samples mixed with myelin, run for technical control.

Table 2.

List of files in the microglia analysis dataset.

Project number PXD066353, DOI: 10.6019/PXD066353
Microglia proteome from the rat substantia nigra. Study in an animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation.
File name/type
PEAK	RAW	RESULT	sample
M01_BA1_1_842.mgf	M01_BA1_1_842_uncalibrated.mzML	M01_BA1_1_842.mzid	SF1
M02_BA2_1_843.mgf	M02_BA2_1_843_uncalibrated.mzML	M02_BA2_1_843.mzid	SF2
M03_BA3_1_844.mgf	M03_BA3_1_844_uncalibrated.mzML	M03_BA3_1_844.mzid	SF3
M04_BA4_1_845.mgf	M04_BA4_1_845_uncalibrated.mzML	M04_BA4_1_845.mzid	SF4
M05_BA5_1_846.mgf	M05_BA5_1_846_uncalibrated.mzML	M05_BA5_1_846.mzid	SF5
M06_BA6_1_847.mgf	M06_BA6_1_847_uncalibrated.mzML	M06_BA6_1_847.mzid	SF6
M07_BA7_1_848.mgf	M07_BA7_1_848_uncalibrated.mzML	M07_BA7_1_848.mzid	SF7
M08_BA8_1_850.mgf	M08_BA8_1_850_uncalibrated.mzML	M08_BA8_1_850.mzid	SS1
M09_BB1_1_851.mgf	M09_BB1_1_851_uncalibrated.mzML	M09_BB1_1_851.mzid	SS2
M10_BB2_1_852.mgf	M10_BB2_1_852_uncalibrated.mzML	M10_BB2_1_852.mzid	SS3
M11_BB3_1_853.mgf	M11_BB3_1_853_uncalibrated.mzML	M11_BB3_1_853.mzid	SS4
M12_BB4_1_855.mgf	M12_BB4_1_855_uncalibrated.mzML	M12_BB4_1_855.mzid	LF1
M13_BB5_1_856.mgf	M13_BB5_1_856_uncalibrated.mzML	M13_BB5_1_856.mzid	LF2
M14_BB6_1_857.mgf	M14_BB6_1_857_uncalibrated.mzML	M14_BB6_1_857.mzid	LF3
M15_BB7_1_858.mgf	M15_BB7_1_858_uncalibrated.mzML	M15_BB7_1_858.mzid	LF4
M16_BB8_1_860.mgf	M16_BB8_1_860_uncalibrated.mzML	M16_BB8_1_860.mzid	LS1
M17_BC1_1_861.mgf	M17_BC1_1_861_uncalibrated.mzML	M17_BC1_1_861.mzid	LS2
M18_BC2_1_862.mgf	M18_BC2_1_862_uncalibrated.mzML	M18_BC2_1_862.mzid	LS3
M19_BC3_1_863.mgf	M19_BC3_1_863_uncalibrated.mzML	M19_BC3_1_863.mzid	LS4

A total of 76 files in the Astrocyte dataset include raw data (*.mzML), peak analysis (*.mgf), results (*.mzid), and search (*.pepXML). Samples SF3, SF5, SF7 are additional samples mixed with myelin, run for technical control.

Table 3.

List of files in astrocyte analysis.

Project number PXD067265, DOI: 10.6019/PXD067265
Astrocyte proteome from the rat substantia nigra. Study in an animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation.
File name/type
PEAK	RAW	RESULT	SEARCH	sample
A01_BA1_1_792.mgf	A01_BA1_1_792_uncalibrated.mzML	A01_BA1_1_792.mzid	A01_BA1_1_792.pepXML	SF1
A02_BA2_1_793.mgf	A02_BA2_1_793_uncalibrated.mzML	A02_BA2_1_793.mzid	A02_BA2_1_793.pepXML	SF2
A03_BA3_1_794.mgf	A03_BA3_1_794_uncalibrated.mzML	A03_BA3_1_794.mzid	A03_BA3_1_794.pepXML	SF3
A04_BA4_1_795.mgf	A04_BA4_1_795_uncalibrated.mzML	A04_BA4_1_795.mzid	A04_BA4_1_795.pepXML	SF4
A05_BA5_1_796.mgf	A05_BA5_1_796_uncalibrated.mzML	A05_BA5_1_796.mzid	A05_BA5_1_796.pepXML	SF5
A06_BA6_1_797.mgf	A06_BA6_1_797_uncalibrated.mzML	A06_BA6_1_797.mzid	A06_BA6_1_797.pepXML	SF6
A07_BA7_1_798.mgf	A07_BA7_1_798_uncalibrated.mzML	A07_BA7_1_798.mzid	A07_BA7_1_798.pepXML	SF7
A08_BA8_1_800.mgf	A08_BA8_1_800_uncalibrated.mzML	A08_BA8_1_800.mzid	A08_BA8_1_800.pepXML	SS1
A09_BB1_1_801.mgf	A09_BB1_1_801_uncalibrated.mzML	A09_BB1_1_801.mzid	A09_BB1_1_801.pepXML	SS2
A10_BB2_1_802.mgf	A10_BB2_1_802_uncalibrated.mzML	A10_BB2_1_802.mzid	A10_BB2_1_802.pepXML	SS3
A11_BB3_1_803.mgf	A11_BB3_1_803_uncalibrated.mzML	A11_BB3_1_803.mzid	A11_BB3_1_803.pepXML	SS4
A12_BB4_1_805.mgf	A12_BB4_1_805_uncalibrated.mzML	A12_BB4_1_805.mzid	A12_BB4_1_805.pepXML	LF1
A13_BB5_1_806.mgf	A13_BB5_1_806_uncalibrated.mzML	A13_BB5_1_806.mzid	A13_BB5_1_806.pepXML	LF2
A14_BB6_1_807.mgf	A14_BB6_1_807_uncalibrated.mzML	A14_BB6_1_807.mzid	A14_BB6_1_807.pepXML	LF3
A15_BB7_1_808.mgf	A15_BB7_1_808_uncalibrated.mzML	A15_BB7_1_808.mzid	A15_BB7_1_808.pepXML	LF4
A16_BB8_1_810.mgf	A16_BB8_1_810_uncalibrated.mzML	A16_BB8_1_810.mzid	A16_BB8_1_810.pepXML	LS1
A17_BC1_1_811.mgf	A17_BC1_1_811_uncalibrated.mzML	A17_BC1_1_811.mzid	A17_BC1_1_811.pepXML	LS2
A18_BC2_1_812.mgf	A18_BC2_1_812_uncalibrated.mzML	A18_BC2_1_812.mzid	A18_BC2_1_812.pepXML	LS3
A19_BC3_1_813.mgf	A19_BC3_1_813_uncalibrated.mzML	A19_BC3_1_813.mzid	A19_BC3_1_813.pepXML	LS4

4. Experimental Design, Materials, and Methods

4.1. Brain surgery

Stereotaxic brain operations were performed under ketamine and medetomidine hydrochloride (intraperitoneally, 75 mg/kg Biowet, Puławy, Poland, and 0.5 mg/kg, Eurovet Animal Health, Holland) anaesthesia. Warm Ringer's solution (1–2 ml) was administered subcutaneously. Eyes were protected (Carbomerum 2 mg/g, Vidisic gel, Dr. Mann Pharma - Bausch Lomb, Germany). Desipramine (30 mg/kg, ip, Sigma-Aldrich, Germany) was administered 30 min before lesioning to protect the noradrenergic terminals. To induce degeneration of dopaminergic neurons the animals were stereotaxically, bilaterally injected with 6-OHDA HBr (3 µg base/3 µl per side), dissolved in 0.2 % ascorbic acid (both from Sigma-Aldrich, Germany) into the passing fibres of the medial forebrain bundle (MFB), at coordinates: AP: 1.4 mm, L: ± 1.6 mm, V: 8.7 mm from bregma, according to Paxinos and Watson`s atlas (2007) (Fig. 1A). Control - sham-operated rats received solvent in the same way. The injection cannula was left in place for 2 min for full absorption of the solution. Additionally, in the same animals, stainless steel cannulas were bilaterally, permanently implanted in the SN pars compacta (SNc) (coordinates: AP: 4.9 mm, L: ± 1.8 mm, V: 8.3 mm from bregma, according to Paxinos and Watson`s atlas (2007)) and connected by a catheter to osmotic minipumps (1007D, ALZET, Charles-Rivers, Germany), implanted under skin on the neck, that administered fluorocitrate (FC, 2 nmol/day, Sigma-Aldrich, Germany) for 7 days, at a continuous rate 0.5 µl / hour, to induce astrocyte dysfunction. Respective control animals had cannulas implanted with sealed catheters. FC was prepared according to Paulsen et al. [5]. The rats received 1 ml of 20 % glucose solution (Pol-Aura, Poland) and meloxicam (Mobic) on the day of operation and 24 hr afterwards. Anaesthesia was reversed with atipamezole hydrochloride (subcutaneously, 1 mg/kg, Orion Pharma, Finland). The body weight of animals was monitored during the whole experiment.

Fig. 1.

Schematic representation of experimental model (A) and analysis workflow (B). Wistar Han rats were bilaterally injected with 6-OHDA (3 µg / 3 µl) into the medial forebrain bundle to induce selective degeneration of dopaminergic neurons and infused for 7 days with fluorocitrate (FC, 2 nmol/day, osmotic minipumps) to induce astrocyte death in the substantia nigra. Substantia nigra tissue was dissected and dissociated into a cell suspension, stained with microglia or astrocyte-specific fluorescent antibodies, and fractionated into astrocytic or microglial clean fractions. Cells were lysed and proteins processed for proteomic analysis. Brain outline based on Paxinos and Watson`s atlas. Created in BioRender. Kuter, K. (2026) https://BioRender.com/1cxstxv.

The animal groups were as follows: SS – sham / sham; SF - sham / FC; LS – 6-OHDA lesion / sham; LF – 6-OHDA lesion / FC. The number of animals per group was 8, resulting in 4 samples per group.

4.2. Brain tissue dissociation

Rats were decapitated on the 7th day after the operation. Both left and right SN were immediately dissected. Tissue from 2 animals was pooled together. The tissue was immediately cooled in HBSS buffer (Gibco™) with 0,2 % glucose. Mechanical and enzymatic dissociation was performed in 37 °C water bath with shaking, in HBSS with glucose, dispase II (Thermofisher), and 0,75 % DNase (Sigma Aldrich), for 20 min, with delicate pipetting. After 10 min, 5 mM Mg²⁺ was added. The process was stopped by dilution with cold HBSS. The digested sample was pushed through an 85 µm mesh, washed, and centrifuged (20 min, 300 x g). Cells were diluted with warm HBSS, 30 % Percoll (Cytiva) added, and centrifuged (30 min, 300 x g). The myelin layer was removed.

4.3. Cell immunofractionation

1 × 10⁶ cells were diluted in cold FACS buffer (DPBS (Gibco™), 1 % BSA, 2 mM EDTA), incubated with Fc blocker (BD Pharmingen™ Purified Mouse Anti-Rat CD32, 10 min) and antibodies (1 hr, Table 4). Cells were washed in a FACS buffer and filtered with a 70 µm mesh.

Table 4.

List of antibodies used.

Antibody	Company, cat. nr
BB700 Mouse Anti-Rat CD45, clone OX-1	BD Biosciences, #742,159
BV421 Mouse Anti-Rat CD11b/c, clone OX-42	BD Biosciences, #743,977
GLAST (ACSA-1) Antibody, anti-human/mouse/rat, APC	Miltenyi Biotec, #130–123–555
Isotype Control Antibody, mouse IgG2a, APC	Miltenyi Biotec, #130–113–831

Using BD FACSAria™ Fusion sorter (BD Bioscience), cells were gated based on FSC-A/SSC-A, excluding debris (Fig. 2). Next, they were gated based on FSC—H/FSC-W to remove doublets and aggregates. The singlet gate was further analysed for the expression of CD11b/c and CD45 to identify microglia (‘Microglia to sort’ gate for sorting), and for GLAST expression to identify astrocytes (‘Astro-like’ gate for sorting). Microglia were distinguished from infiltrating peripheral macrophages based on the expression level of CD45. The population with low expression of CD45 was defined as microglia [6]. For isotype control, mouse IgG2a, APC antibody (Miltenyi Biotec, cat. 130–113–831) was used. FACS analysis and sorting were performed with the following controls: fluorescence compensation (CompBeads Anti-Mouse Ig, κ Negative Control Compensation Particles Set, BD Biosciences #552,843), isotype control (mouse antibody IgG2a, APC, 1:400, Miltenyi Biotec, #130–113–831), positive control with the peritoneal cavity cells stained with specific antibodies, and negative control with unstained brain cells. Microglia cells were identified and sorted out as CD11b/c+ and CD45+, while astrocytes were identified as GLAST⁺. Cell sorting purity was 92,2 % for microglia and 55 % for astrocytes. Cells were collected in SDS buffer with protease inhibitors, sonicated, and stored in −80 °C until further use. Raw data on FACS sorting are available in Supplementary Materials.

Fig. 2.

Sorting strategy of astrocytes stained with GLAST and microglia stained with both CD11b/c and CD45, dissociated from adult brain substantia nigra using BD FACSAria™ Fusion sorter. Microglia were differentiated from circulating macrophages by CD45 expression. Schematic Created in BioRender. Kuter, K. (2026) https://BioRender.com/2e9g09e using original research files showing control sample from sham+solvent group.

4.4. Cell culture and validation

Cells after tissue dissociation (Fig. 3A) and after FACS sorting out astrocytes (Fig. 3B) and microglia (Fig. 2C) were plated in Lab-Tek® Chamber Slide (Merck, #C6932) covered with poly-d-lysine in DMEM/F12 medium with GlutaMax (Gibco, #10,565,018), 20 % FBS (Thermo Fisher, #26,140,079), Pen/Strep (Sigma Aldrich, #P4333). Culture in standard conditions (37 °C, 5 % CO₂) was continued up to 4 weeks.

Fig. 3.

Phase contrast microphotographs (Leica DM IL LED) of cells cultured for 4 weeks after adult rat brain tissue dissociation (A) and after FACS sorting out GLAST⁺ astrocytes (B) and CD11b/c⁺CD45⁺ microglia (C). 20 x magnification.

4.5. MS analysis and protein identification

Mass spectrometry: The protein extracts were reduced in 10 mM TCEP at 60 °C for 30 min., alkylated by 40 mM iodoacetamide for 30 min. at room temperature and finally digested by trypsin (Promega, Trypsin Gold, Mass Spectrometry Grade) applying sp3 method on amine magnetic beads 20 µg/µl (MagReSyn) in KingFisher Flex instrument (Thermo Scientific) at 50 °C for 4 h. Then, samples were acidified by the addition of 10 % formic acid and were analysed on the timsTOF Pro 2 (Bruker) instrument performing DDA (data-dependent acquisition) proteomics, operating in positive-ion mode. Peptides obtained by digestion were separated on Ultimate 3000 RSLC nano system (Thermo Scientific) with bioZen C18 nano column 250 × 0.075 mm (Phenomenex) with trap cartridge Acclaim PepMap C18 5 × 0.3 mm (Thermo Scientific) using 60 min. gradient from 2 % to 35 % of acetonitrile with 0.1 % (v/v) formic acid. The mass spectrometer was equipped with an ion mobility separation (IMS) feature, allowing the analysis of ions in a 1/K0 range of 0.6 to 1.6. Data were acquired over an m/z range of 100 to 1700, and the collected data were processed for protein identification and quantification. The DDA settings of the mass spectrometer for PASEF (parallel accumulation serial fragmentation) scan were the default.

The DDA data were processed by FragPipe version 19.0 software using MS Fragger 3.6, IonQuant 1.8.9, and Philosopher 4.7.0. Search was performed against a FASTA file containing a proteome of Rattus norvegicus (Rat), including all reviewed (Swiss-Prot) and unreviewed (TrEMBL) proteins (47,921), proteome ID UP000002494. Search parameters were set as follows: modifications: N-term acetylation, Met oxidation, and Cys carbamidomethylation, up to 2 missed cleavages; peptide length range: 5–50 amino acids, all other parameters were as default for trapped ion mobility spectrometry (TIMS) with PASEF data. Identified proteins were statistically processed in Perseus software version 2.0.10.0 (Max Planck Institute). All data were normalized based on the average median.

4.6. Data analysis

Qualitative data analysis was performed using Venn diagrams (Python 3.11.7 packaged by Anaconda, Inc., and Jupyter Notebook version 7.0.8) to illustrate sets of proteins with lost (zero) expression and those with new (positive) expression. Protein expression was analysed in paired treatment groups.

Semi-quantitative analysis was performed using Volcano Plots (Perseus 2.0.11). Raw data were uploaded as *.tsv files. Log2 transformation was applied. Histograms were used to analyse the distribution shape of protein intensities in each sample. For normalization ``Subtract'' method was applied. Rows were filtered based on a minimum number of valid values (≥ 3). Proteins had to be detected in at least 3 out of 4 samples per group. Volcano plot was performed using multiple hypothesis testing (t-test, log2 fold change, FDR = 0.05; s₀ = 0.1). Results were exported from Perseus and visualized in Python 3.11.7. Fold change thresholds were set from −0.8 to +0.8. A final list of differentially expressed proteins between groups was compiled.

Functional analysis was performed using Functional Annotation Bioinformatics Microarray Analysis (Database for Annotation, Visualization, and Integrated Discovery (DAVID), v2025_1, https://davidbioinformatics.nih.gov/; NIH) and GO Enrichment categories. Identified proteins were referred to according to their gene names. Gene lists derived from both qualitative and semi-quantitative analyses were submitted to the DAVID database for the R. norvegicus species. The output tables included Gene Ontology (GO) terms, gene counts, and fold enrichment values. Results were visualized with Python.

Pathway analysis was performed using the KEGG Pathway Database (https://www.genome.jp/kegg/pathway.html) via the DAVID web interface.

Protein-Protein Interaction Networks were visualized as a network graph using the STRING – Functional Protein Interaction Network database (https://string-db.org/).

Validation criteria

Validation of methods used was summarised in Table 5.

Table 5.

Validation criteria for methods used in this research.

Animal model	Brain injection coordinates were experimentally tested on an animal cohort and adapted to animal size. Injection precision was visually tested during tissue dissection. Quality of fluid pumping by osmotic minipumps was verified after a priming period and prior to cannula implantation by visual inspection of fluid passed through the catheters. Volume of FC pumped into SN was checked by control of residual fluid in the pump after explantation. Each set of osmotic minipumps had been certified by Alzet pumping rate with standard deviation. Any outlier animals or tissue were excluded. Exclusion criteria for tissue collection were brain haemorrhage, abscess, and weight loss. Only verified tissues were pulled for further processing.
Cell sorting	BD FACS Aria Fusion was used. Method parameters were extensively validated before the experiment. Fluorescence compensation was performed with BD™ CompBead particles. Other controls included isotype control antibodies, a positive control with the peritoneal cavity cells stained with specific antibodies, negative control with unstained brain cells. BD Fc Block was used to reduce nonspecific binding of antibodies to Fc receptors. Additional validation was performed by in vitro culture of dissociated cells, as well as sorted cells and staining with antibodies. BD FACSDiva 9.0.1 software was used. To further check the purity of the collected population, a small number of the sorted cells was re-analysed on a flow cytometer.
Protein extraction	The known number of cells per µl was verified by protein quantitation test before MS analysis and pilot studies.
Proteomic analysis	The timsTOF Pro 2 mass spectrometer was calibrated using ESI-L Low Concentration Tuning Mix (Agilent Technologies, Cat. No: G1969–85,000). Before each sample set the instrument was additionally calibrated using ESI nano source and lock masses (for mobility and mass calibration): m/z 622.028961: 1 µg/µl in isopropanol CAS number: 186,817–57–2; Hexakis(2,2-difluoroethoxy)phosphazene (Apollo Scientific, Cat. No: PC1165); m/z 922.009799: 1 µg/µl in isopropanol CAS number:58,943–98–9; Hexakis(2,2,3,3-tetrafluoropropoxy)phosphazene (Apollo Scientific, Cat. No: PC0874) and m/z 1221.990637: 2 Vol % in isopropanol CAS number: 186,406–47–3; Chip cube high mass reference (HP-1221) (Agilent Technologies, Part No: G1982–85,001). The instrument was operated by timsControl software version 6.0.8.
Data analysis	Each protein must have been detected in 3 out of 4 samples in order to be taken to further analysis. FDR (False Discovery Rate) was set as <5 %. Log₂ fold-change thresholds were set at ≤ −0.8 and ≥ 0.8. In accordance with the p ≤ 0.05, –log₁₀(p-value) ≥ 1.3. Python 3.11.7 packaged by Anaconda, Inc., and Jupyter Notebook version 7.0.8. and Perseus 2.0.11.

Limitations

Substantia nigra is a very small brain structure, approximately 1 mm³ per side of the brain. In order to obtain enough cells sorted, bilateral structures from 2 animals were pooled together. This always generates data spread. Furthermore, dissociation and isolation of good-quality cells from adult rat brains is a huge challenge. Astrocytes are arbored cells with very small cell bodies, easily damaged during dissociation; therefore, it was difficult to discriminate them from cell debris in FACS analysis, hence a lower fraction purity than from microglia. Another limitation could be the sample size. Pooling tissues allowed us to generate 4 samples per group. Therefore, the data really come from n = 8 animals per group. Having a bigger sample size, probably more of significant changes could have been detected.

Ethics Statement

Three-month-old male Wistar HAN rats (Charles Rivers, Germany) were kept under a 12-hour dark/light cycle (light from 06:00 to 18:00), with free access to food and water. The experiments were carried out in compliance with the Animal Experiments Bill of January 21, 2005; (published in Journal of Laws no 33/2005 item 289, Poland), and according to the EU Directive 2010/63/EU for animal experiments; as well as National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No 8023, revised 1978). They also received approval from the Local Ethics Committee (approval no: 221/2018; 221A/2018, 306/2018; 33/2020; 242/2020, 259/2022). All efforts were made to minimize the number of animals and their suffering. Experiments complied with the ARRIVE guidelines.

CRediT Author Statement

K.Z. Kuter: Conceptualization, Funding acquisition, Methodology, Investigation, Validation, Formal analysis, Writing - Original Draft, Writing - Review & Editing, Supervision, Project administration; J. Kadłuczka, T. Chubukova, P. Mielczarek, A. Maziak, A. Roman, E. Napieralska, J. Kula: Methodology, Investigation, Formal analysis, Data Curation, Writing - Review & Editing.

Data Availability

• ProteomeXchangeMicroglia proteome from the rat substantia nigra. Study in animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation (Original data).

• ProteomeXchangeAstrocyte proteome from the rat substantia nigra. Study in animal model of fluorocitrate-induced astrocyte dysfunction and microglia activation (Original data).