Blood samples were obtained following acquisition of the study participants’ informed consent. The protocol was approved by ethic committees of San Camillo Forlanini Hospital. Patients enrolled were diagnosed with relapsing-remitting form of MS according to 2005-revised McDonald’s criteria (34). Frontal cortex tissue was collected postmortem by UK MS Tissue Bank at Imperial College, London, and procedures for retrieval, processing, and storage have gained ethical approval.

Peripheral blood mononuclear cells were isolated by a density gradient centrifugation over a Ficoll-Hypaque (Ficoll-Paque PLUS, GE Healthcare) from 20 ml of freshly venous blood from five healthy donors (HD), five relapsing MS patients (MS acute), and five remitting MS patients (MS stable). Cells were stained with pre-titrated Abs, to evaluate the expression of P2X7R within cluster of differentiation 14 (CD14)-positive cells. Briefly, PBMCs (1 × 10⁶) were incubated with P2X7-extracellular epitope antibody (Alomone Labs, Jerusalem, Israel) for 30 min at 4°C. Cells were washed and stained with goat anti-rabbit Alexa Fluor 488-conjugated antibody (Invitrogen, Life Technologies, Monza, MB, Italy), 30 min at 4°C. Cells were washed and stained with anti-CD14 PE (Dako, Aurogene, Rome, Italy) and Live Dead Fixable Aqua Dead Cell Stain Kit (Invitrogen) for 30 min at 4°C.

Monocytes were isolated from PBMCs of HD by using Magnetic Separation with Negative Selection Columns (Miltenyi Biotec, Calderara di Reno, BO, Italy) according to the product manual. Purified monocytes (6 × 10⁶) were cultured in serum-free RPMI 1640 with L-Glutamine, 50 U/ml penicillin, 50 µg/ml streptomycin in 96-well plates.

FACS analysis was performed with FACS CyAn (Beckman Coulter, Pasadena, CA, USA) and with advanced flow cytometry software FlowJo (Tree Star, Ashaland, OR, USA).

Human and rat monocytes or snap-frozen tissues were homogenized in TRIzol (Life Technologies) and total RNA was extracted following the manufacturer’s instructions. UV spectrophotometric analysis of nucleic acids was performed by Nanodrop spectrophotometer at 260 nm to determine concentration. The 260:280 absorbance ratio was used to assess nucleic acids purity. After DNase treatment (Qiagen, Hilden, Germany), 1 µg of total RNA was subjected to retro-transcription by high-capacity RNA-to-cDNA kit (Applied Biosystem, Life Technologies).

Quantitative polymerase chain reaction was carried out using SYBR green (Applied Biosystem, Life Technologies) incorporation with gene-specific primers (Table 1). Relative gene expression was calculated by ΔΔCt analysis relative to GAPDH.

Female Lewis rats (~160 g, 6 weeks old) were purchased from Charles River (Como, Italy). Animal procedures were performed according to European Guidelines for animal use in research (86/609/CEE) and requirements of Italian laws (D.L. 116/92), according to protocol no. 112/2014B by Italian Ministry of Health. Efforts were made to minimize animal suffering and the number of animals used.

Female rats were deeply anesthetized and injected in each hind paw with 100 µl of a medium containing 0.15 g/ml guinea pig spinal cord tissue in saline (0.9% NaCl) and complete Freund’s adjuvant (CFA, Sigma-Aldrich, Milan, Italy), 50% vol/vol, to which 5 mg/ml heat-inactivated Mycobacterium tuberculosis (Difco H37Ra) were added. CFA-injected rats were used as control of inflammation.

Starting at 5-day postinjection, all animals were daily weighed, assessed for clinical signs of disease, and graded according to the following described criteria: 0 = no clinical signs; 1 = loss of tail tonus; 2 = weakness in one or both hind legs or mild paresis; 3 = severe paresis or paralysis of both hind legs; 4 = severe paralysis of complete lower part of the body; and 5 = death due to aggressive EAE (35).

Female C57BL/6 mice (~25 g, 8 weeks old) were purchased from Charles River (Como, Italy). Animal procedures were performed according to European Guidelines for animal use in research (86/609/CEE) and requirements of Italian laws (D.L. 116/92), according to protocol no. 119/2015PR by Italian Ministry of Health. Efforts were made to minimize animal suffering and the number of animals used.

Female Lewis rats (n = 4) and C57BL/6 mice (n = 3) were sacrificed by asphyxiation with CO₂ and spleen excised for monocytes purification. CFA (n = 3) and EAE (n = 4) rats were sacrificed at 15 days postinjection by asphyxiation with CO₂. After spleen harvest, single cell suspension was obtained by mechanical tissue dissociation in RCB buffer (NH4Cl 150 mM, NaHCO3 10 mM, and EDTA 1 mM). Cells were plated (4 × 10⁶/ml) in RPMI, 10% fetal bovine serum, 100 U/ml gentamycin, 100 µg/ml streptomycin, and 100 U/ml penicillin. After 2 h, non-adherent cells were removed and medium enriched with 10 ng/ml rat or mouse macrophage colony-stimulating factor (Sigma-Aldrich) (36). After 1 week, cells were used for RT-qPCR, western blotting, and immunofluorescence analysis.

Human monocytes were stimulated without or with lipopolysaccharide (LPS, Sigma-Aldrich) or 2′(3′)-O-(4-Benzoylbenzoyl)adenosine 5′-triphosphate (BzATP, Sigma-Aldrich) for 4 h (T4) and 24 h (T24) at 37°C in a 5% CO₂ environment. After treatments, monocytes were incubated with monoclonal antibody as described above, to evaluate P2X7R expression by FACS analysis. Moreover, monocytes from control rats and mice were stimulated in vitro with or without LPS or BzATP for 4 h (T4) or 24 h (T24) and P2X7R expression was analyzed by western blotting.

Demographic and clinical characteristics of MS cases at the time of tissue collection are reported (Table 2). Frontal cortex tissues are from 13 neuropathological confirmed cases of MS, matched for disease courses (all secondary progressive MS, SPMS) but presenting different ages (range 34–80 years), sex, disease durations (range 6–50 years), and causes of death (see Table 2). Analysis was performed also on samples from four patients who died by non-neurological diseases. Cerebral hemispheres were fixed with 4% paraformaldehyde for 2 weeks, coronally sliced, and blocked. Individual blocks were cryoprotected in 30% sucrose for 1 week, frozen in isopentane, and stored at −80°C until use.

Immunohistochemistry was performed as described (37). Human sections (30–40 µm) were pre-incubated for 10 min with 5% H2O2 in 5% methanol in PBS, and for 24–48 h in PBS-0.3% Triton X-100, 2% normal donkey serum (NDS) at 4°C, with goat anti-P2X7 receptor antibody (1:100, peptide YETNKVTRIQSMNY-C from the N-terminus of human P2RX7 corresponding to amino acid residues 13-26, MyBioSource, San Diego, CA, USA). Sections were then incubated with biotinylated donkey anti-goat antibodies (Jackson ImmunoResearch Europe Ltd., Suffolk, UK), followed by avidin–biotin–peroxidase reactions (Vectastain, ABC kit, Vector, Burlingame, CA, USA), using 3,3′-diaminobenzidine (Sigma-Aldrich) as a chromogen. Some sections were counterstained with Luxol fast blue, in order to detect lesion types. Sections were mounted on poly-lysine slides and air dried for 24 h. The histological preparations were examined using an Axioskop 2 light microscope (Zeiss). Images were taken with Neurolucida software (MBF Bioscience, USA).

Human sections (30–40 µm) were blocked with 10% NDS in 0.3% Triton X-100 in PBS and incubated with primary antisera/antibodies (Table 3) in 0.3% Triton X-100 and 2% NDS in PBS, for 24–48 h at 4°C and processed for double and triple immunofluorescence. The secondary antibodies in 0.3% Triton X-100 and 2% NDS in PBS were Cy3-conjugated donkey anti-goat IgG (1:100, Jackson Immunoresearch, West Grove, PA, USA, red), Cy5-conjugated donkey anti-mouse IgG (1:100, Jackson Immunoresearch, blue), Alexa Fluor^® 488-AffiniPure donkey anti-mouse IgG (1:200, Jackson Immunoresearch, green), and Alexa Fluor^® 488-AffiniPure donkey anti-rabbit IgG (1:200, Jackson Immunoresearch, green). In the case of biotinylated primary antibody CD14 and Lectin from Lycopersicon esculentum (tomato) biotin conjugate (1:200, Sigma-Aldrich), Cy2-streptavidin conjugated secondary antibodies (1:200, Invitrogen) were used.

Rat monocytes maintained in culture for 1 week were fixed with 4% paraformaldehyde for 20 min, permeabilized with 0.1% Triton X-100 for 10 min, blocked in PBS/1% bovine serum albumin, and incubated in PBS/1% bovine serum albumin with anti-cluster of differentiation 68 (CD68, 1:200, Santa Cruz Biotechnology, CA, USA). The secondary antibody was Cy3-conjugated donkey anti-rabbit IgG (1:200, Jackson Immunoresearch). Cells were stained with Höechst 33342 (1:1,000) and covered with coverslip in Fluoromount medium (Sigma-Aldrich) for confocal microscopy.

Quantification of CD68 immunoreactivity was performed with monocytes from rat spleen (from an average of six fields for each animal in each group), using Image J software. Data are expressed as optical density with respect to CFA group.

Immunofluorescence analysis was performed by confocal laser scanning microscope (Zeiss, LSM700; Iena, Germany) equipped with four laser lines: 405, 488, 561, and 639 nm. Brightness and contrast were adjusted with Zen software (Zeiss).

Six different snap-frozen blocks of frontal cortex from three independent SPMS cases, and one block from five different control cases were processed for protein extraction. Detergent-soluble proteins were extracted with Ripa buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS in PBS, containing protease inhibitors), using a micropestle. After a short sonication, the homogenates were incubated on ice for 1 h and centrifuged at 13,000 rpm for 10 min at 4°C. To extract detergent-insoluble proteins, the resulting pellet was resuspended in 0.5 M Tris HCl (pH 6.8) containing 2% w/v SDS (same volume as utilized for soluble protein extraction), and incubated at RT for 10 min. Insolubilized material in the pellet (15,000 rpm for 15 min) was discarded. Protein quantification was performed from the supernatants by Bradford colorimetric assay (Biorad, Milan, Italy). Proteins (15 µg) were separated by electrophoresis on 10% SDS-PAGE and transferred to nitrocellulose Hybond-C-extra membranes (Amersham Biosciences, Cologno Monzese, Italy). The filters were pre-wetted in 5% blocking agent in TBS-T (10 mM Tris pH 8, 150 mM NaCl, 0.1% Tween 20) and hybridized overnight with P2X7-extracellular epitope antibody (1:500, peptide KKGWMDPQSKGIQTGRC, corresponding to amino acids 136–152 of mouse P2X7 receptor, Alomone Labs), in the absence or presence of the neutralizing immunogenic peptide used in a 1:1 protein ratio, and with monocyte chemoattractant protein-1 (MCP-1) antibody (1:1,000, ThermoFisher Scientific). The signals were detected with anti-rabbit horseradish peroxidase-conjugated antibody (1: 5,000), and developed by enhanced chemiluminescence (Amersham Biosciences), using Kodak Image Station (KDS IS440CF) and semi-quantitative analysis of which was performed with Image J software.

Total proteins from control (n = 4), CFA (n = 3), and EAE (n = 4) cultured rat monocytes and from control C57BL/6 cultured mouse monocytes (n = 3) were obtained after lysis of cells on ice for 1 h in Ripa Buffer and centrifugation at 14,000 rpm for 10 min at 4°C. Protein quantification was performed from the supernatants by Pierce™ BCA Protein Assay Kit (Thermo Scientific, Monza, MB, Italy). Proteins were separated by electrophoresis on 10% SDS-PAGE, transferred to nitrocellulose membranes, and processed for western blotting using P2X7-intracellular epitope antibody [1:500, peptide (C)KIRKEFPKTQGQYSGFKYPY, corresponding to amino acids 576–595 of rat P2X7 receptor, Alomone Labs]. The signal was detected with anti-rabbit horseradish peroxidase-conjugated antibody (1: 2,500), and developed by enhanced chemiluminescence (Amersham Biosciences). The bands of interest were visualized using Kodak Image Station (KDS IS440CF) and semi-quantitative analysis was performed with Image J software.

Data are presented as mean ± SEM and statistical analysis was determined by ANOVA. Statistical differences between groups were verified by Student’s t-test. *p < 0.05, **p < 0.01, and ****p < 0.0001 were considered statistically significant.

Given the abundant expression of P2X7R on monocytes (38–40), in this work we firstly confirmed its presence in circulating monocytes from HD and then demonstrated its down-regulation in stable and acute MS patients compared with HD, by RT-qPCR (Figure 1A). Furthermore, FACS analysis confirmed a significant decrease of CD14/P2X7R-positive monocytes only in MS acute patients compared with stable MS and HD conditions (Figure 1B).

Because the rat model of EAE is one among the most commonly used animal model for studying MS pathogenesis, by resembling particularly the acute form of the disease (41), we observed also in EAE monocytes, compared with CFA, a statistically significant decrease of P2X7R mRNA by RT-qPCR (Figure S1A in Supplementary Material), and of protein content by western blot analysis (Figure S1B in Supplementary Material). This occurs in parallel to the increase of the specific monocyte inflammatory marker IL-6 mRNA (Figure S1C in Supplementary Material), and of CD68 protein (Figure S1D in Supplementary Material).

In order to mimic an inflammatory insult as it occurs in MS, monocytes from HD were challenged ex vivo with LPS or BzATP for 4 and 24 h. A significant decrease of human P2X7R/CD14 bearing monocytes occurred after treatment for 24 h with LPS (Figures 1C,E, ~48% reduction) or BzATP (Figures 1D,E, ~63% reduction), as shown by FACS analysis. Western blot analysis confirmed these results in vitro in rat cultured monocytes, showing down-regulation of P2X7R protein after 4 and 24 h of BzATP (Figure S1E in Supplementary Material, ~60 and 90% reduction, respectively) or LPS stimulation for 24 h (Figure S1E in Supplementary Material, ~45% reduction). Similar results were also obtained with mouse purified monocytes challenged in vitro with LPS or with BzATP. (Figure S1F in Supplementary Material).

By immunohistochemistry and immunofluorescence analysis, we next analyzed the cortical tissue from 13 different cases of SPMS for the presence of P2X7R immunoreactivity and colocalization with specific cellular markers (Tables 2 and 3). In detail, we examined 1–2 different tissue blocks from all cases and inspected 4–10 different slices for each block. The tissue slices were studied in areas either presenting neuronal injury/inflammation, or devoid of visible damage. We observed typical features of cortical demyelination and inflammation in all SPMS cases analyzed. Independent analysis was also performed in cortical tissue from four patients who died by non-neurological diseases (data not shown). In particular, by immunohistochemistry (Figure 2A) the P2X7R immune-positive signal was found to delineate the plasma membrane of roundish cells distributed in distinct clusters in the cortical parenchyma. By immunofluorescence analysis, this roundish P2X7R immune-positive signal (red) was found to colocalize with the CD45 leukocyte marker (Figure 2B, green). By staining with Lectin (green) that specifically delineates endothelial vascular cells, we thus concluded that the roundish P2X7R-positive cluster cells were likely located inside blood vessels (Figures 2C,D) within the SPMS cortical parenchyma. Similar results were confirmed in all cases analyzed (Table 2) and in donors not deceased by neurological conditions (data not shown). By performing double and triple confocal immunofluorescence of these cluster cells with the CD14 monocyte/macrophage marker (Figures 2E–G, green), we demonstrated only partial colocalization with P2X7R immunoreactivity (Figures 2E–G, in red), being the P2X7R signal apparently absent from CD14-positive cells (white arrows) that are proximal to the endothelium of the blood vessels and that are simultaneously positive for CD68 (Figure 2F, blue, white arrows) or for the major histocompatibility complex II (MHC II) (Figure 2G, blue, white arrows) macrophage/microglia markers. In all the sections analyzed, P2X7R is present only on few perivascular double CD14- and MHC II-positive cells (Figure 2G, arrowheads, white signal).

We next investigated the distribution of P2X7R in the cortical parenchyma outside from the blood vessels. Double immunofluorescence confocal analysis indicated the absence of colocalization of P2X7R (red) with P2Y12R (green) or MHC II (blue) (markers, respectively, of quiescent or reactive/active macrophages/microglia, Figures 3A,B) (42, 43), or with myelin basic protein (MBP) (blue, marker of myelin fibers, and oligodendrocyte cell bodies, Figure 3C). The receptor was instead strongly expressed in glial fibrillary acidic protein (GFAP)-positive astrocytes present in both gray (GM, Figures 3D–F) and white matter (WM) (not shown) of both control and MS patients. Moreover, immunohistochemistry studies corroborated the presence of P2X7R in the soma and fibers of interlaminar astrocytes (Figure 3G) present in GM cortical layer I that spread prominent, long, and unbranched processes throughout the layers of the cortex, and moreover in protoplasmic astrocytes (Figure 3H) that are well-organized in GM into non-overlapping spatial domains. In WM, we also observed a strong P2X7R signal in fibrous astrocytes (Figure 3I) that exhibit unbranched cellular processes and that often protrude “vascular feet” (44) that are physically connected to the external capillary walls (Figure 3J).

With the aim of further characterizing P2X7R expression and modulation in astrocytes, we acquired images from adjacent immunohistochemical fields within the same cortical sections and compared them with control cases (Figures 4A–C). Respect to control (Figure 4A), we observed a remarkable increase in P2X7R-positive astrocytes in different WM zones of SPMS cortical tissue presenting high levels of astrogliosis (Figure 4B) and glial scar formation (Figure 4C). In detail, in SPMS cortex, P2X7R immunoreactivity distinguished an area (Figure 4B) with strong reactive astrogliosis, intense cellularity, prominent hypertrophy, proliferation, and overlapping of astrocyte processes causing the disruption of distinctive astrocyte domains. Furthermore, the P2X7R signal also identified a zone very rich in parallel and interconnected fibers highlighting a prominent glial scar, where astrocytes displayed densely intersecting processes characterized by intense double P2X7R/GFAP-positive immunoreactive signal (Figure 4C; Figure S2 in Supplementary Material).

Although MS is still widely considered a WM disease, demyelination is also prominent in GM. SPMS phase is characterized by subpial lesions that are highly abundant in progressive stages of MS, closely to the subarachnoid space, involving either part of a cortical gyrus, or often covering adjacent gyri (45). By triple immunofluorescence confocal analysis, we have characterized a subpial lesion with various degrees of inflammatory reaction and demyelination and observed different features of P2X7-positive immunoreactivity. In detail, the presence of profuse reactive MHC II-positive monocytes/macrophages/microglia (blue, Figure 5B) and the permanence of MBP-positive myelin fibers (green, Figure 5C) defined a chronic active lesion characterized by abundant P2X7R-positive signal (red, Figure 5A and insets) that highlighted a zone of intense cellularity and astrogliosis (see merged P2X7R-GFAP signal in the inset). On the other hand, in an area where MHC II (blue, Figure 5E) and MBP (green, Figure 5F) signals both decreased indicating a chronic inactive lesion (46), the P2X7R-positive immunoreactivity (red, Figure 5D) identified a zone very rich in fibers typical of a glial scar (see merged P2X7R-GFAP signal in the inset).

To further investigate if SPMS progression modifies the P2X7R content, we analyzed total cell extracts from frontal cortex, by RT-qPCR and western blotting. A statistically significant increase of P2X7R mRNA was observed in SPMS patients, compared with controls (Figure 6A). Similar results were confirmed by immunoblotting with a P2X7R antibody raised against an extracellular epitope of the mouse P2X7R (corresponding to amino acid 136–152). In human frontal cortex tissue, we recognized three specific protein bands with estimated sizes in the ranges 52–72, 72–95, and 95–140 kDa, perhaps corresponding to the different P2X7 isoforms, and that were moreover abolished by the P2X7R neutralizing immunogenic peptide (data not shown). In addition, particularly the 95–140 kDa P2X7R was found significantly up-regulated in both detergent-soluble and -insoluble fractions of SPMS tissue extracts respect to controls, while the 52–72 kDa P2X7R was significantly up-regulated only in the detergent-insoluble fraction of SPMS extracts, with a trend to increase in the detergent-soluble fraction (Figures 6B,C).

Because the up-regulation of MCP-1 in astrocytes is demonstrated to have an important role in recruiting leukocytes in the CNS during MS (47–49), and the P2X7R agonist BzATP increases MCP-1 expression in astrocytes through P2X7R activation (50), we evaluated the expression of this chemokine in our human SPMS cortical tissue and its potential colocalization with P2X7R on astrocytes. Western blot analysis of protein extracts from control and SPMS patients, demonstrated a strong up-regulation (about twofold increase) of MCP-1 in MS respect to control (Figure 7). Moreover, triple immunofluorescence confocal analysis showed an unambiguous intense colocalization (white signal) among P2X7R (red), MCP-1 (green), and GFAP (blue) proteins, both in control (Figures 8A–D) and SPMS tissues (Figures 8E–H), and moreover a strong up-regulation of both P2X7R and MCP-1 during astrogliosis occurring in WM of SPMS patients (Figures 8E–H).