Purinergic receptors P2Y12R and P2X7R: potential targets for PET imaging of microglia phenotypes in multiple sclerosis

Chemicals were obtained from commercial sources and used without further purification. Solvents were purchased from Sigma-Aldrich (Zwijndrecht, The Netherlands), MERCK (Darmstadt, Germany) and Biosolve (Valkenswaard, The Netherlands) and used as received unless stated otherwise. Tetrahydrofuran (THF) was first distilled from LiAlH₄ and then stored on 3 Å molecular sieves. [³H]methyl nosylate stock solution (717 MBq/mL in toluene, molar radioactivity of 3.08 GBq.μmol⁻¹) was obtained from Novandi Chemistry AB (Södertälje, Sweden) and dried at 60 °C under an argon flow prior to use.

Brain tissue from three controls and eight MS patients were used in this study. Patient details are presented in Additional file 1: Table S1. Brain tissue samples were obtained from the Netherlands Brain Bank (coordinator Dr. Huitinga, Amsterdam, The Netherlands). The Netherlands Brain Bank received permission to perform autopsies for the use of tissue and for access to medical records for research purposes from the Ethical Committee of the VU University Medical Center, Amsterdam, The Netherlands. All patients and controls, or their next of kin, had given informed consent for autopsy and use of brain tissue for research purposes.

We used EAE tissues acquired from an independent study performed in our laboratory and the acute experimental autoimmune encephalomyelitis (EAE) was induced as follows. Eight- to 11-week-old male Lewis rats (200 g) obtained from Harlan (Zeist, The Netherlands), as described before [25]. Rats were injected s.c. with 20 μg synthetic myelin basic protein 63–88 peptide, 500 μg Mycobacterium tuberculosis type 37HRa (Difco, Detroit, MI, USA), and 50 μl complete Freund’s adjuvant (CFA) (Difco) supplemented with PBS to reach a volume of 100 μl. Rats were examined daily (weight and clinical disease) and graded on a scale from 1 to 5 for neurological signs. Clinical disease in EAE animals was apparent around day 10 post-immunization with a maximum clinical score between days 14 and 15. Animals were housed under standard laboratory conditions with water and food ad libitum. CFA-immunized animals were used as a control. Animals were sacrificed at day 14 dpi (peak of the disease) and day 20 dpi (recovery phase, end of clinical signs).

For human tissue staining, air-dried frozen sections (5 μm) were fixed with acetone (10 min at RT). For single immunohistochemistry staining of proteolipid protein (PLP), MHC-II, purinergic receptor P2X7R, and P2Y12R, we used the protocol described previously by Vogel et al. [4] using Envision-HRP (Dako) and 3,3′-diaminobenzidine (Dako) as detection method.

For immunofluorescence staining on human tissue, sections were fixed with acetone then blocked for non-specific binding with goat serum (10%) for 20 min at RT. Sections were then incubated with the primary antibody (anti-P2Y12R or anti-P2X7R) in PBS/1% serum overnight at 4 °C. Sections were then washed with PBS three times for 5 min and incubated with fluorescence-labeled secondary antibody (1/400) in PBS/1% serum or for 1 h at RT. Sections were then washed three times for 5 min with PBS and blocked with mouse serum (10%) for 20 min at RT followed by incubation with anti-MHC-II or anti-CD31 or anti-GFAP antibody in PBS/1% BSA for 1 h at RT. After three times of washes for 5 min, a fluorescently labeled secondary antibody was added in PBS/1% BSA for 1 h at room temperature. Nuclei were stained with Hoechst (1/1000) for 1 min in the dark. After a final wash for three times for 5 min with PBS, sections were mounted with coverslips using aqueous mounting media Mowiol.

For rat tissue immunofluorescence double staining, air-dried frozen sections (7 μm) were used. Tissues from three different EAE animals at day 14 post-immunization and three different EAE animals at day 20 post-immunization, and three CFA control rat brains were used. Sections were fixed with acetone for 10 min at RT and blocked for non-specific binding using BSA 2%. Sections were then incubated with a mix of both primary antibodies overnight at 4 °C or 1 h at room temperature. Sections were then washed three times in PBS/0.05% Tween-20 and then incubated with fluorescently labeled secondary antibodies and DAPI (1/1000) for nuclear staining for 1 h at RT. Sections were then washed three times for 5 min with PBS/0.05% Tween-20. Sections were mounted with coverslips using aqueous mounting media Mowiol.

Fluorescence images were acquired using Leica DM6000 microscope equipped with a motorized stage. Whole-slide low-zoom images were acquired using ×10 objective and merged by tile stitching to get the final image. For fluorescence quantification and comparison, sections were stained in a single run and images were collected in a single session with the same exposure time between different areas and different slides. Bright field images were acquired on a Zeiss bright field microscope equipped with a colored camera. Images were analyzed using Leica LAS AF or Zen 2012 software. The list of antibodies used for immunohistochemistry and immunofluorescence are described in Additional file 1: Table S3.

Human microglia isolation and preparation was performed as described elsewhere [3]. In brief, 5 to 10 g of brain white matter was obtained at autopsy, and microglia isolation procedure was performed within 4 to 24 h. Single-cell suspension was prepared using 0.05% trypsin (Sigma, St Louis, MO). Cells were then filtered through a 100-μm nylon mesh (BD Bioscience, Durham, NC) and centrifuged, and the pellet is resuspended in a gradient buffer (3.56 g/L Na₂HPO₄·2H₂O, 0.78 g/L NaH₂PO₄·H₂O, 8 g/L NaCl, 4 g/L KCl, 2 g/L d-(+)-glucose, and 2 g/L BSA, pH 7.4) and centrifuged for 35 min at 1200×g. The cell debri-myelin layer was removed. The cell pellet was treated with red blood cell lysis buffer (8.3 g/L NH₄CL and 1 g/L KHCO₃, pH 7.4) for 10 min at 4 °C. Cells were then suspended and cultured in DMEM/Ham nutrient mixture F10 (1:1), 1% penicillin-streptomycin-glutamine, and 25 ng/mL granulocyte-macrophage colony-stimulating factor (GM-CSF is only added for the first 2 days). After 7 days, microglia were polarized into a pro-inflammatory status by priming with recombinant IFN-γ (1 × 103 U/mL) (U-Cytech, Utrecht, The Netherlands) for 24 h, followed by addition of Escherichia coli LPS (10 ng/mL) (LPS-EB ultrapure; Invitrogen, San Diego, CA) to the medium for 24 h. To induce anti-inflammatory microglia, cells were stimulated with IL-4 (10 ng/mL) (Immunotools, Friesoythe, Germany) for 48 h. Untreated cells were referred to as M0.

All oligonucleotides were synthesized by Ocimum Biosolutions (Ocimum Biosolutions, Ijsselstein, The Netherlands). RNA was isolated using triazol. cDNA was synthesized with the Reverse Transcription System kit (Promega, Madison, WI, USA) following manufacturer’s guidelines as described previously [26]. QPCR reactions were performed on an Applied Biosystems ViiA 7 machine with the SYBR Green method (Applied Biosystems, Carlsbad, CA, USA). Obtained expression levels of transcripts were normalized to GAPDH and PolRF2 expression levels. Here, we report only the results normalized to GAPDH as we did not observe any difference of normalized data between GAPDH and PolRF2. All primer sequences are listed in Additional file 1: Table S2.

Cells were lysed with RIPA lysis buffer containing protease inhibitors for 1 h on ice. The cells were then centrifuged at 14,500 rpm for 15 min, the pellet was discarded, and the supernatant was collected. Electrophoresis was performed under denaturating conditions on 10% SDS polyacrylamide gel. The proteins were transferred to a nitrocellulose membrane. Blots were saturated in Odyssey buffer/PBS (1:1) for 1 h at RT. The blots were incubated with primary antibodies Rabbit anti-human P2X7R (Alomone, Jerusalem BioPark (JBP), Israel), or Rabbit anti-human P2Y12R (Ananspec, LIEGE Science Park, Belgium) in Odyssey buffer/PBS (1:1) + Tween-20 (0.1%) overnight at 4 °C followed by fluorescently labeled secondary antibody (Odyssey) in Odyssey buffer/PBS (1:1) + Tween-20 (0.1%) for 1 h at room temperature. Blots were imaged using Odyssey imager. Blot were then washed and stained with goat anti-human β-actin (Santa Cruz, Heidelberg, Germany) antibody following the same procedure as above. Western blot relative quantification was realized by calculating the ratio of the integrated intensity of the band of the protein of interest over the integrated intensity of the band of β-actin.

For autoradiography tissues from three non-neurological controls, three active, two chronic active, and three chronic inactive MS lesions, and from three different EAE animals at day 14 post-immunization and three different EAE animals at day 20 post-immunization, and three CFA control rat brains were used. Tissue cryosections (20 μm) of acute EAE and human MS were mounted on glass slides, air dried and stored at −80 °C until use. Sections were washed in assay buffer (50 mM Tris-HCl, pH 7.4) for three times 5 min and then dried under an air flow. Sections were incubated with [¹¹C]P2Y12R-ant (10 nM) for 30 min at room temperature, or incubated with [³H]A740003 for 1 h and 30 min at room temperature. Sections were then washed three times for 90 s with Tris-HCl (50 mM, pH 7.4) and dipped twice in deionized water. Sections were dried under an air flow and exposed to phosphor screen BAS-IP TR 2025 (for tritum-labeled tracer) or phosphor screen BAS-IP SR 2025 (for carbon-11-labeled tracer) (General Electric, Eindhoven The Netherlands) for 1 h for [¹¹C]P2Y12R or 48 h for [³H]A740003. Phosphor screen were then imaged using Typhoon FLA 7000 imager (General Electric, Eindhoven, The Netherlands). The intensity of the signal was quantified using image Quant software (General Electric, Eindhoven, The Netherlands). Three independent experiments were performed.

All data are presented as mean ± SD. Groups were compared using the two-tailed Student t test (normal), and Mann-Whitney (non-normal) and analysis were done with Graphpad prism 5 (San Diego, CA, USA). p values of less than 0.05 were considered statistically significant.

To investigate the regulation of P2Y12R and P2X7R expression, human microglia were freshly isolated from 10 different adult post-mortem brains, cultured for 7 days and then polarized into an inflammatory phenotype with IFN-γ/LPS or anti-inflammatory phenotype with IL-4 as described previously [3] (Additional file 1: Figure S1). Expression levels of P2Y12R, P2X7R, and TSPO were measured using quantitative PCR. P2Y12R was significantly downregulated in inflammatory polarized microglia compared to non-stimulated microglia (fourfolds lower, p < 0.001). In anti-inflammatory polarized microglia, the expression of P2Y12R was strongly increased compared to non-stimulated microglia (threefolds higher, p < 0.01) and pro-inflammatory polarized microglia (12-folds higher, p < 0.001) (Fig. 1a).

P2X7R expression revealed an opposite expression pattern with a significant increase in pro-inflammatory microglia (fivefolds higher, p < 0.01) compared to non-stimulated microglia and anti-inflammatory stimulated microglia (fivefolds higher, p < 0.01). No significant change was observed in anti-inflammatory microglia, which indicates a strong association of P2X7R with a pro-inflammatory microglia phenotype (Fig. 1b). The increased expression of P2X7R in pro-inflammatory microglia and P2Y12R in anti-inflammatory microglia was also observed at the protein level as shown by western blot from the same donors (Fig. 1d, e). Protein level was only evaluated in two donors due to the low amount of microglia cells obtained from the isolations which makes it challenging to have enough protein for western blot analysis. We next investigated the expression of TSPO as this marker is considered to be the state of art for PET imaging of activated microglia. TSPO was highly expressed in pro-inflammatory microglia (1.5-folds higher, p < 0.05), albeit less pronounced compared to P2X7R (1.5-folds vs fivefolds, respectively). In anti-inflammatory microglia, no significant change of TSPO expression was observed compared to non-stimulated microglia (Fig. 1c).

In order to investigate P2Y12R and P2X7R as targets for imaging microglia phenotype in MS, we analyzed their expression in white matter samples from three non-neurological controls, three active, two chronic active, and three chronic inactive MS lesions (control, Fig. 2a–d; active, Fig. 2e–l; chronic active, Fig. 2m–p, chronic inactive, Fig. 2q–t; Additional file 1: Figure S2-5). P2X7R and P2Y12R were expressed on ramified cells in control brain and normal-appearing white matter (NAWM) that were identified as microglia by their colocalization with MHC-II (Fig. 3). In inflamed MS lesions, P2X7R and P2Y12R are expressed on activated microglia characterized by their round shape and high MHC-II expression (Figs. 2 and 3; Additional file 1: Figure S2-S5). GFAP- (marking activated astrocytes) and CD31- (marking endothelial cells) positive cells in MS lesion, NAWM, or control tissue did not express P2X7R or P2Y12R (Additional file 1: Figure S7-10).

P2X7R and P2Y12R expression on microglia is affected by their activated state. To compare the change of expression of P2X7R and P2Y12R between ramified and activated microglia, we performed colocalization studies with MHC-II in active and chronic active MS lesions. P2X7R was more robustly expressed on activated microglia in active lesions and on microglia in the active rim of the chronic active lesions compared to microglia in the NAWM (Fig. 3m–x; Additional file 1: Figure S11). In contrast, P2Y12R immunolabeling showed a gradual decrease of expression on microglia from the NAWM towards the center of the active lesion (Fig. 3a–l). Interestingly, at the border of the active lesion, we observe a mixed population of activated microglia expressing either high or low levels of P2Y12R; the same pattern was also observed in activated microglia in the rim of the chronic active lesions (Additional file 1: Figure S12).

We observe that P2X7R and P2Y12R are differentially expressed within MS active lesions. In order to investigate the dynamic of expression of P2X7R and P2Y12R, we compared their respective change of expression on ramified microglia in the NAWM and activated microglia in MS lesions. Enhanced expression of P2X7R on activated microglia in the MS lesion is accompanied by a decrease in microglial P2Y12R expression on the same cells (Fig. 4a, b). To further evaluate the relation between the expression of P2Y12R and P2X7R, we quantified their expression by drawing regions of interest (ROIs) around microglia in the NAWM and MS lesion. When comparing the microglia at the ramified less activated state to the highly activated state, we detected that in the majority of microglia an increase of expression of P2X7R coincides with a decrease of P2Y12R expression (Fig. 4c, d). We also observed a small population of activated microglia that shows an increase of P2X7R expression and no decrease in P2Y12R expression; these cells were mostly located at the border of the lesion.

To investigate whether the change of expression of P2Y12R and P2X7R that we observed in post-mortem human MS tissue also occurs in EAE, a well-established animal model of MS, we performed immunohistochemistry on tissues from acute EAE induced in Lewis rats. We used cerebellum and brain stem tissue sections where the majority of neuroinflammation and immune cell infiltration occurs in this particular model. Tissues were collected from EAE animals at the peak of the disease (day 14 post-immunization) and in the recovery phase (day 20 post-immunization). Tissues from animals immunized with complete Freund’s adjuvant (CFA) were used as control. Microglia were stained with anti-CD11b, and sections were analyzed by differentiating a highly inflamed and a less inflamed area based on the CD11b staining.

We first investigated the expression of P2X7R and P2Y12R in control animals. P2Y12R immunoreactivity localized with CD11b-positive microglia in control brain tissue (Additional file 1: Figure S13). No P2X7R expression was detected in control brain tissue, which may be due to the low expression of P2X7R and/or the low sensitivity of the antibody (Additional file 1: Figure S16).

At the peak of the disease, we observed a marked decrease of P2Y12R on CD11b-positive cells in the highly inflamed area. Quantification of the P2Y12R fluorescence signal revealed a significant reduction of the staining in the inflamed area compared to the less inflamed area (Fig. 5a–d). In EAE, at the end of the recovery phase, P2Y12R is markedly expressed on CD11b-positive microglia compared to microglia in the less inflamed area (Fig. 6a, b). The quantification of the P2Y12R expression showed a significant increase of the P2Y12R expression in the inflamed area compared to the less inflamed area (Fig. 6c, d). Infiltrating monocyte-derived macrophages (ED1-positive) in the perivascular area did not express P2Y12R (Additional file 1: Figure S14).

P2X7R was highly expressed on CD11b-positive cells in the EAE tissue at the peak of the disease and at the end of the recovery phase. The expression of P2X7R on activated microglia in the inflamed area was higher compared to the ramified microglia in the less inflamed area (Fig. 7a, b).

To investigate the binding of PET tracers and their ability to predict and follow the change of P2Y12R and P2X7R expression, we performed autoradiography binding experiment using a carbon-11-labeled specific P2Y12R tracer ([¹¹C]P2Y12R-ant) and tritium-labeled P2X7R-specific tracer ([³H]A-740003) for binding to EAE and human MS post-mortem tissues. For these binding experiments, we used sections from the same human MS and EAE tissues blocks as for the immunostaining experiments. In tissue from acute EAE animals, we observed a significant decrease of binding of the [¹¹C]P2Y12R-ant tracer at the peak of the disease compared to the control that increases back in the recovery phase. P2X7R tracer, [³H]A-740003, showed a significant increase of binding to EAE tissues from the peak of the disease compared to control. The binding of [³H]A-740003 slightly decreased in the recovery phase (Fig. 8a–c). In human MS tissues, the binding of [¹¹C]P2Y12R-ant tracer was significantly decreased in all types of MS lesions (active, chronic active, and chronic inactive) compared to control (Fig. 8d, f). The binding of P2X7R tracer [³H]A-740003 was significantly increased only in the MS active lesion while there was a slight non-significant increase in chronic active lesions and a decrease in the chronic inactive lesions compared to control (Fig. 8e, f). We also performed blocking studies and post-autoradiography immunostaining; our data showed that the binding of both tracers is specific and matches the antibody staining (Additional file 1: Figure S17-19). The binding of the P2X7R and P2Y12R tracers correlated with our immunohistochemistry results which further validate these receptors as potential targets for use in PET imaging of activated microglia.