Fc gamma receptors are expressed in the developing rat brain and activate downstream signaling molecules upon cross-linking with immune complex

The in vivo and in vitro experiments were conducted at the developmental times illustrated in the schematic below:

For the majority of studies, which were performed at UC Davis (Davis, CA), timed pregnant female Sprague Dawley rats were purchased from Charles River Laboratories and housed individually in standard plastic cages with Alpha-Dri bedding (Shepherd Specialty Papers) in a temperature-controlled room (22 ± 2 °C) on a normal 12 h light/dark cycle. Food and water were provided ad libitum.

For the surfaceome analyses, which were performed at the ETH (Zurich, Switzerland), Sprague Dawley rats were purchased from Charles River Laboratories and group housed (three per cage) in Allentown individually ventilated cages with Lignocel bedding (JRS - Rettenmaier) in a temperature-controlled room (21 ± 2 °C) on a normal 12-h light/dark cycle. Food and water (Kliba) were provided ad libitum.

Primary cultures of dissociated hippocampal or cortical cells were prepared and maintained as previously described [45]. Briefly, hippocampi and neocortices from pooled postnatal day 0 (P0) or P1 male and female rat pups were dissociated and plated onto tissue culture plastic or on glass coverslips precoated with poly-L-lysine (Sigma), at different densities depending on the endpoint to be analyzed (detailed information provided in the respective method section). Cells were maintained in Neurobasal A media supplemented with B27 and 2 mM Glutamax (all from Life Technologies), except for cultures used in the proteomics experiments, for which cells were maintained in Neural Q supplemented with GS21 (both from Sigma). Cytosine-D-arabinofuranoside (araC; 5 μM, Sigma) was added to the medium on day in vitro (DIV) 4 to prevent overgrowth of astrocytes. Half the medium was replaced with fresh media (without araC) twice weekly.

Resident peritoneal macrophages were collected from adult female Sprague Dawley rats, plated at a density of 2 × 10⁵ cells/cm² and maintained as described [46]. Cells were plated on tissue culture plastic (Thermo Fisher Scientific) and allowed to adhere for 1 h at 37 °C in a 5% CO₂ humidified incubator. Non-adherent cells were removed by washing the plates twice with warm RPMI 1640 medium (Gibco). This procedure enriches remaining adherent cells for macrophages (~ 95%) [47]. Using this macrophage-enriched population of peritoneal cells, specificity of the anti-FcγRI (anti-CD64) and anti-FcγRII (anti-CD32) antibodies was determined by confirming lack of CD64/CD32 immunolabeling on CD3+ cells and positive CD64/CD32 immunolabeling on CD3- cells.

To assess Fcgr transcript levels in vivo, rat pups were euthanized on P1 or P7 (n = 3 pups of each sex and age from the first litter; n = 3 female and 3 male pups at P1 and 3 female and 2 male pups at P7 from the second litter). Sex was determined by anogenital distance. P1 pups were euthanized by decapitation; P7 pups were placed on a cloth-lined Petri dish on top of ice to induce hypothermic anesthesia, and then decapitated, and liver, spleen, and brain were removed for further analysis. To assess in vitro Fcgr expression, cultured hippocampal or cortical cells were harvested from P1 male and female pups (n = 6 cultures per experimental group with 3 cultures each from 2 independent dissections; cells from both sexes were pooled). Immediately following euthanasia, the brain (microdissected into hippocampus, cortex, and cerebellum, after carefully removing meninges), liver, and spleen were harvested and placed in RNAlater overnight (Life Technologies). The following day, RNAlater was removed and the samples stored at − 80 °C until further analysis. To assess in vitro expression of Fcgr, cell lysates were collected from hippocampal and cortical neuron-glia co-cultures (seeded at a density of 0.5–1 × 10⁵ cells/cm²) on DIV 2, 7, 14, and 21. To investigate the effects of IFNγ on Fcgr transcript expression, cortical cell cultures were exposed to recombinant rat IFNγ (100 ng/ml, Peprotech) added to the culture medium on DIV 1 for 24 or 48 h (3 cultures derived from each of 6 independent dissections were used). Cells were lysed in Buffer RLT (RNeasy Mini Kit, Qiagen), lysates homogenized using QIAShredder spin columns (Qiagen), and homogenates stored at − 80 °C until further processed.

Extraction of total RNA and synthesis of cDNA from brain tissues or cell cultures were performed as previously described [48]. cDNA was loaded in duplicate in MicroAmp Optical 384-well plates (Life Technologies), and qPCR was performed at the Real-Time PCR Research and Diagnostics Core Facility at UC Davis using Taqman technology on a 7900HT Fast Real-Time PCR System (Applied Biosystems). Primer and probe sets, designed to span an exon-exon junction, were obtained from Integrated DNA Technologies (IDT). Gene accession numbers and primer/probe set product number and amplification efficiencies are presented in Table  1 .

Specific amplification of products of the desired length was verified using gel electrophoresis (not shown). qPCR results were analyzed using the Applied Biosystems 7900HT Fast Real-Time PCR System software (SDS, Thermo Fisher Scientific; version 2.4) to obtain Ct values [49]. Across all qPCR experiments, each sample was quantified in duplicate, and duplicate Ct values were averaged for analysis of results.

Relative differences in gene expression between experimental groups were calculated as described previously [48]. Transcript levels were normalized to the concentration of the reference gene Pgk1 [48, 50], and relative expression ratios were calculated by the Pfaffl method [51] using REST 2009 software (Qiagen). Consistent with prior studies demonstrating that Pgk1 is a stably expressed gene in the brain [50, 52], expression levels of Pgk1 were not different between experimental conditions.

Cortical and hippocampal cell cultures (seeded at a cell density of 1 × 10⁵ cells/cm²) were lysed in ice-cold lysis buffer [50 mM Tris, pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.1% sodium dodecyl sulfate (SDS), 0.5% sodium deoxycholate, and 1X Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific)]. The protein concentration of cell lysates was determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific), and 30 μg of protein from each sample was separated by SDS PAGE using 10% Bolt Bis-Tris Plus Gels (Invitrogen), transferred onto PVDF membranes using the iBlot 2 dry transfer system, and immunoblotted using the iBind Western System (Life Technologies). Proteins were identified based on expected molecular weight of immunoreactive bands as indicated in the product information provided by the supplier from which antibodies were purchased: 40–45 kDa for FcγR isoforms (FcγRIa: rabbit polyclonal anti-CD64, Santa Cruz Biotechnology sc-15364, RRID AB_2103451, 10 μg/ml; FcγRIIIa: rabbit monoclonal anti-CD16, Abcam ab109223, RRID AB_10863447, dilution 0.5 μg/ml), 37 kDa for GAPDH (rabbit monoclonal anti-GAPDH, Cell Signaling Technology 2118, RRID AB_561053, dilution 1:6000) and 55 kDa for βIII-tubulin (mouse monoclonal anti-β-III-tubulin, Millipore MAB1637, RRID AB_2210524, dilution 1:6000). GAPDH (37 kDa) was used as a loading control for the western blots of FcγRIa (45–50 kDa); βΙΙΙ-tubulin (55 kDa), for western blots of FcγRIIIa (37 kDa). Since all tissues express FcγR, and we could not locate an FcγR knockout mouse, loading buffer was used as a negative control; lysates of peritoneal macrophages (30 μg per lane) were used as a positive control. Densitometric analysis was performed using Image Studio Lite software (RRID SCR_014211), v4.0 (Licor Biotechnology). Densitometric values for target proteins were normalized to densitometric values of loading controls (GAPDH for FcγRIa and βIII-tubulin for FcγRIIb, FcγRIIIa, and FcRn) in the same sample. FcγR protein expression was investigated in 2 cultures from each of 2 independent dissections.

Rat cortical cells were dissociated, suspended in phosphate-buffered saline (PBS) containing 2% fetal bovine serum (FBS, LifeLine Cell Technology), and divided into flow cytometry tubes (5 ml polystyrene round-bottom tubes) for labeling, with 5 × 10⁶ cells/ml/tube. Cells were pelleted by centrifugation at 250×g for 5 min and re-suspended in primary antibody solutions: anti-FcγRI (rabbit polyclonal anti-CD64, Santa Cruz Biotechnology sc-15364, RRID AB_2103451, 2 μg/ml) and anti-FcγRII (mouse anti-rat CD32, BD Biosciences 550271, RRID AB_393568, 2 μg/ml) in PBS with 2% FBS. Cells were incubated for 40 min at room temperature on an orbital shaker. Cells were then centrifuged and washed with PBS, fixed for 30 min with 2% paraformaldehyde (PFA, Sigma), and permeabilized with 0.25% Triton X-100 (Fisher Scientific) in PBS for 10 min. After permeabilization, cells were centrifuged and incubated with primary antibodies against MAP2b (guinea pig polyclonal anti-MAP2b, Synaptic Systems 188,004, RRID AB_2138181, dilution 1:1000) or GFAP (guinea pig polyclonal anti-GFAP, Synaptic Systems 173,004, RRID AB 10641162, dilution 1:1000) in PBS containing 1% bovine serum albumin (BSA) and 10% goat serum for 1 h at room temperature. Cells were washed twice by adding PBS to each tube, shaking vigorously and then centrifuging for 5 min before re-suspending in PBS. Cells were incubated with fluorescently labeled secondary antibodies in PBS for 45 min at room temperature. Phycoerythrin (PE)-labeled secondary antibody (PE-conjugated donkey anti-guinea pig F(ab’)2, Jackson ImmunoResearch Labs 706-116-148, RRID AB_2340455, dilution 1:1000) was used to detect MAP2b- and GFAP-labeled cells. AlexaFluo-488 goat anti-mouse F(ab’)2 antibody was used to detect FcγRII (CD32)-labeled cells (Jackson ImmunoResearch Labs 115-546-006, RRID AB_2338860, dilution 1:1000), while AlexaFluor 647 goat anti-rabbit IgG (H+L) was used to detect FcγRI (CD64)-labeled cells (Thermo Fisher Scientific A-21245, RRID AB_2535813, dilution 1:1000). After incubation with secondary antibodies, the cells were washed twice with PBS and kept at 4 °C until analysis. A subset of cells were fixed and incubated without antibodies as a negative control. Additional controls included cells incubated with one primary antibody at a time, and cells incubated only with secondary antibodies alone and in combination, in order to set the maximum threshold for non-specific binding. A FACS Calibur (BD Biosciences) instrument was used for acquisition and compensation at the UC Davis Flow Cytometry Shared Resource. Compensation was performed using the data from cells incubated with each of the cellular markers individually. Data were analyzed and plotted using FlowJo V10 (RRID SCR_008520, FlowJo, LLC, Ashland, OR).

For cell surface staining, cortical and hippocampal cell cultures were grown on glass coverslips (5–9 × 10⁴ cells/cm²) and fixed on DIV 7 in 4% PFA for 20 min at room temperature. Cells were blocked with PBS containing 1% BSA and 10% goat normal serum (Vector Laboratories) for 1 h at room temperature. After blocking, cultures were stained with primary antibodies specific for FcγRI (rabbit polyclonal anti-CD64, Santa Cruz Biotechnology sc-15364, RRID AB_2103451, 2 μg/ml) or FcγRII (mouse anti-rat CD32, BD Biosciences 550271, RRID AB_393568, 2 μg/ml) in PBS with 2% FBS overnight at room temperature. Cultures were then rinsed with PBS, permeabilized with 0.25% Triton X-100 in PBS for 5 min and stained with antibodies specific for GFAP (guinea pig polyclonal anti-GFAP, Synaptic Systems 173 004, RRID AB_10641162, dilution 1:1000) and MAP2b (guinea pig polyclonal anti-MAP2b, Synaptic Systems 188 004, RRID AB_2138181, dilution 1:1000) for 1 h at room temperature. After incubation with primary antibodies, cultures were rinsed 3 times with PBS and then incubated with secondary antibodies (AlexaFluor-568 goat anti-guinea pig IgG, Thermo Fisher Scientific A-11075, RRID AB_141954; AlexaFluor-647 goat anti-rabbit IgG, Thermo Fisher Scientific A-21245, RRID AB_2535813; and AlexaFluor-488 goat anti-mouse F(ab’)2, Jackson ImmunoResearch Labs 115-546-006, RRID AB_2338860, all diluted 1:1000) for 30 min. Cells were then rinsed twice with PBS, quickly rinsed in water, and mounted on slides in Prolong Gold Antifade with DAPI (Molecular Probes, Thermo Fisher Scientific). Images were acquired using a 40× objective on the ImageXpress MicroXL high-content screening system (Molecular Devices).

Cell surface capture (CSC) technology was used as previously described [53] to identify the cell surface proteotype (e.g., surfaceome) in DIV 7, DIV 14, and DIV 21 hippocampal cell cultures (seeded at a density of 1 × 10⁵ cells/cm²). Briefly, surface glycoproteins were gently oxidized with 2 mM NaIO4 and subsequently biotinylated with 5 mM biocytin hydrazide for 20 min. Cells were harvested by scraping and lysed in lysis buffer (50 mM ammonium bicarbonate, 0.25% Rapigest, 10 mM TCEP) by sonication using a VialTweeter (Hielscher Ultrasonics Gmbh, Germany). Proteins were alkylated for 30 min with 10 mM iodoacetamide (room temperature in the dark) and digested with trypsin overnight at 37 °C, in an enzyme-to-protein ratio of 1:50. In order to inactivate trypsin, samples were boiled 20 min and subsequently incubated with 1 mM PMSF for 15 min at 37 °C while shaking (300 rpm) and then acidified with 10% formic acid to pH 3. Samples were centrifuged (10 min, 16,000 × g) before capturing biotinylated glycopeptides using streptavidin affinity matrix (Pierce, Ultralink). Glycopeptides captured on the streptavidin affinity matrix were washed as previously described and incubated with PNGase F at 37 °C for 18 h to release N-glycosylated peptides. Samples were acidified to pH 2–3 with formic acid, and peptides were desalted with C18 UltraMicroSpin columns (The Nest Group) according to the manufacturer’s instruction and dried in a SpeedVac concentrator (Thermo Scientific, Rockford, USA).

For mass spectrometric (MS) analysis, peptides were reconstituted in 3% acetonitrile and 0.1% formic acid, and separated by reverse-phase chromatography on a high-pressure liquid chromatography (HPLC) column (75-μm inner diameter; New Objective) packed in-house with a 15-cm stationary phase (Magic C18AQ, 200 Å, 1.9; Michrom Bioresources) and connected to an EASY-nLC 1000 instrument combined with an autosampler (Proxeon). The HPLC was coupled to a Q Exactive plus mass spectrometer equipped with a nanoelectrospray ion source (Thermo Scientific). Peptides were loaded onto the column with 100% buffer A (99% H2O, 0.1% formic acid) and eluted at a constant flow rate of 300 nL/min for 30 min with a linear gradient from 6 to 20% buffer B (99.9% acetonitrile, 0.1% formic acid), followed by 15 min from 20 to 35% buffer B and a washing step with 90% buffer B. Mass spectra were acquired in a data-dependent manner (top 15). In a standard method for medium-to-low abundant samples, high-resolution MS1 spectra were acquired at 70,000 resolution (automatic gain control target value 3 × 10⁶) to monitor peptide ions in the mass range of 375–1500 m/z, followed by high-energy collisional dissociation (HCD)-MS/MS scans @ 35,000 resolution (automatic gain control target value 1 × 10⁶). To avoid multiple scans of dominant ions, the precursor ion masses of scanned ions are dynamically excluded from MS/MS analysis for 30 s. Single-charged ions and ions with unassigned charge states or charge states above 6 were excluded from MS/MS fragmentation.

For data analysis, SEQUEST software v27.0 was used to search fragment ion spectra for a match to tryptic peptides with maximally two missed cleavage sites from a protein database, which was composed of the Rattus norvegicus reference proteome (Uniprot, organism ID 10116), various common contaminants and sequence-reversed decoy proteins. The precursor ion mass tolerance was set to 20 ppm. Carbamidomethylation was set as a fixed modification on all cysteines. The PeptideProphet and the ProteinProphet tools of the trans-proteomic pipeline (TPP v4.6.2) were used for the probability scoring of peptide spectrum matches, and inferring protein identifications were filtered to reach an estimated false-discovery rate of ≤ 1%. Finally, proteins were filtered for cell surface glycoproteins as previously described [53]. In brief, each identified protein was required to be identified with at least one peptide containing the consensus NXS/T glycosylation motif containing an asparagine to aspartic acid deamidation site with a MS measured mass difference of 0.986 Da, indicating the cell surface labeling and the enzymatic deamidation during the CSC workflow.

To form an antibody-antigen immune complex, normal mouse IgG and affinity-purified rat anti-mouse IgG with minimal cross-reactivity to rat serum proteins were used as previously described [54]. The mouse IgG was dialyzed overnight against sterile PBS using a 10,000 Da molecular weight cutoff Slide-A-Lyzer MINI dialysis device (Thermo Fisher Scientific) to remove sodium azide. The dialysate was spin-concentrated using Amicon ultra centrifugal filter units (Millipore) before complexing. The IgG-IC was prepared fresh before each experiment by incubating mouse IgG with rat anti-mouse IgG in a 1:1 ratio (w/w) for 1 h at room temperature prior to addition to the cells, with the exception of the internalization studies, where incubation of cultures with each component of the immune complex (AlexaFluor 647-labeled rat anti-mouse IgG, Jackson ImmunoResearch Labs 415-605-166, RRID AB_2340285; and normal mouse IgG, Santa Cruz Biotechnology sc-2025, RRID AB_737182, concentration ratio 1:1) was performed sequentially. Cultures treated with the individual components of the IgG-IC were used as controls.

To measure intracellular calcium concentration ([Ca²⁺]_i), dissociated hippocampal neurons were seeded at a density of 1.5–3.0 × 10⁴ cells/cm² (1–3 wells per condition in 4 independent dissections) and cultured on Costar 96-well plates (Corning). On DIV 7, cells were loaded with the Ca²⁺-sensitive dye, Fluo-4 AM (2 μM; Thermo Fisher Scientific), in imaging buffer (10 mM HEPES 140, pH 7.4, mM NaCl, 5 mM KCl, 2 mM MgCl₂, 2 mM CaCl₂, and 10 mM glucose) for 30 min in a 5% CO₂ humidified incubator. Cultures were then washed with imaging buffer and transferred onto the stage of the ImageXpress MicroXL high-content imaging system. IgG-IC was applied to the cells, and full-frame FITC images were captured in one randomly chosen field per treatment (2 frames/s for 2.0–2.2 min) using MetaXpress software (Molecular Devices). Image capture was initiated at t = 0 s; a half volume of IgG-IC or of a rat anti-mouse IgG control or a mouse IgG control was added to an equal volume of imaging media at t = 10 s (20th frame); KCl (50 mM) was added at the end of each experiment (t = 110 s) to verify cell viability. A total of 240–260 images were acquired per well, and 1–3 wells were used per treatment in cultures from 4 independent dissections. Images were acquired in sequential mode between individual wells so that each well would receive the respective treatment (vehicle or IgG-IC) and images would be acquired from that well before proceeding to treatment of the next well. In instances where more than one well per condition were used, images from one well per condition across conditions were acquired as a first set, then the whole procedure was repeated for the next set of wells, in order to minimize as much as possible the influence of time between incremental exposures. Microscope stage coordinates were recorded for each chosen field.

To quantify [Ca²⁺]_i, images were thresholded to remove background fluorescence and regions of interest (ROIs) were drawn around the soma of each Fluo-4-loaded cell. Fluorescence intensity values across all acquired images per ROI were logged into Excel spreadsheets (Microsoft Corporation). Custom Excel macros were written to automate data analysis. Briefly, ROIs with an area < 30 μm² were excluded from analysis, and fluorescence data (F) for each of the remaining ROIs were normalized to its baseline fluorescence value (F ₀ ; frames 1–20). Only cells in which addition of KCl resulted in a spike in normalized fluorescence intensity with amplitude greater than 10% were included in further analyses. For each cell, the maximum fluorescence following application of IgG IC (F; measured in arbitrary fluorescence units) was normalized to the baseline fluorescence (F ₀ ) for that cell and expressed as (F − F₀)/F₀ or ΔF/F₀. Cells were considered responsive to treatment when the peak change in fluorescence intensity (ΔF/F ₀ ) after treatment was > 50% of baseline fluorescence. The total number of IgG-IC-responsive cells in 1–3 wells for each concentration of IgG-IC was divided by the total number of cells in those wells to derive the percentage of the cells responding. Data from IgG-exposed cells were compared to data from cells exposed to an equal volume of fresh imaging media.

DIV 7 hippocampal cell cultures were treated with fresh culture media, IgG-IC at 10 or 100 μg/ml, or rat anti-mouse IgG at 10 or 100 μg/ml (1 culture per condition in each of 3 independent dissections). Following 30 min of exposure at 37 °C in a 5% CO₂ humidified incubator, cultures were lysed with ice-cold lysis buffer (as described for western blotting). Protease and phosphatase inhibitors (1X Halt Protease and Phosphatase Inhibitor Cocktail; Thermo Fisher Scientific) were added to the lysis buffer immediately before cell lysis. Protein concentration was determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific), and 30 μg of protein from each sample was separated using 7.5 or 10% Mini-PROTEAN TGX precast gels (Bio-Rad), then transferred to PVDF membrane. Membranes were blocked with Odyssey Blocking Buffer (Licor Biotechnology) for 30 min at room temperature and then incubated overnight at 4 °C with primary antibodies against phospho-Erk1/2 (Phospho Thr 202 & Tyr204/rabbit monoclonal anti-phospho-p44/42 MAPK (Erk1/2), Cell Signaling Technology 9106, RRID AB_331768, dilution 1:1000) diluted in blocking buffer with 0.1% Tween-20. Antibodies against GAPDH were used as loading controls. After washing in PBS supplemented with 0.1% Tween-20, membranes were incubated with secondary antibodies diluted in blocking buffer with 0.1% Tween-20 and 0.01% SDS for 1 h at room temperature. After extensive rinsing, membranes were scanned using the Odyssey Infrared Imaging System (Licor Biotechnology). Blots were then incubated with stripping buffer (62.5 mM Tris-HCl, pH 6.8, with 2% SDS and 100 mM β-mercaptoethanol) for 1 h at 37 °C and then imaged to confirm the absence of bands. Stripped membranes were immunoblotted for total (phosphorylated and non-phosphorylated) Erk1/2 (Erk 1/2/mouse monoclonal anti-p44/42 MAP kinase, Cell Signaling Technology 4696, RRID AB_390780, dilution 1:2000). Antibodies against GAPDH (rabbit monoclonal anti-GAPDH, Cell Signaling Technology 2118, RRID AB_561053, dilution 1:6000) were used as loading controls. Following incubation with secondary antibodies, membranes were scanned using the Odyssey Infrared Imaging System (Licor Biotechnology). Secondary antibodies used for these experiments were IR700DX goat anti-mouse IgG (Rockland 610-130-121, RRID AB_220121), IR700DX goat anti-rabbit IgG (Rockland 611-130-122, RRID AB_220148), IR800 goat anti-mouse IgG (Rockland 610-132-121, RRID: AB220125) and IR800 goat anti-rabbit IgG (Rockland 611-132-122, RRID AB_220152), all diluted 1:10000. Densitometry values of bands, immunoreactive with antibody against total Erk1/2 (phosphorylated and non-phosphorylated) or antibody that specifically recognizes phosphorylated Erk, were normalized to densitometry values for GADPH-immunopositive bands in the same sample. Normalized data for treated samples were expressed as a percentage of normalized data for vehicle controls from the same experiment. As a positive control, lysates were collected from peritoneal macrophages (one culture per condition in a single experiment) exposed to 10 or 100 μg/ml IgG-IC for 5 min at 37 °C in a 5% CO₂ humidified incubator.

To test whether binding of an immune complex (AlexaFluor 647-labeled rat anti-mouse IgG and normal mouse IgG) to FcγR leads to internalization of FcγR into endosomes, AlexaFluor 488-conjugated rat transferrin (Jackson ImmunoResearch Labs 012-540-050, RRID AB_2337161) was used to label the entire endocytic recycling pathway [55]. Briefly, hippocampal cell cultures (seeded at a density of 5 × 10⁴ cells/cm²) were exposed on DIV 7 to fresh growth media or 100 μg/ml of rat anti-mouse IgG for 10 min, followed by incubation with a mixture of 100 μg/ml of normal mouse IgG and 30 μg/ml labeled transferrin for 10 min. Incubations were performed at 37 °C in a 5% CO₂ humidified incubator. Negative controls were incubated with rat anti-mouse IgG and transferrin, but without normal mouse IgG. In a subset of experiments, cultures were pre-incubated for 5 min with anti-FcγRI (Santa Cruz Biotechnology sc-15364, RRID AB_2103451) or anti-FcγRIIb (BD Biosciences 550271, RRID AB_393568) antibodies (at 10 μg/ml each) to block the FcγR.

Following the 10-min transferrin exposure, cells were fixed with 4% PFA. Images were acquired at × 40 magnification using the ImageXpress MicroXL high-content imaging system. Twenty-five sites were imaged per well in two wells per condition in each of 3 (without Fc block) or 5 (with Fc block) independent dissections. Images were analyzed using MetaXpress software. A custom data analysis module was created using Custom Module Editor to calculate the average number of rat anti-mouse fluorescent puncta co-localized with transferrin per cell.

Graphs were constructed in Graphpad Prism (RRID SCR_002798; v5.01), and statistical analyses were performed using SPSS (RRID SCR_002865; version 22.0.0.1) unless otherwise specified. Statistical comparisons were performed using block design ANOVA (experimental days were used as blocks; a Tukey’s honest significant difference/HSD test was used to identify significant differences between groups) with the following exceptions: (i) comparisons of Fcgr transcript levels between hippocampal and cortical cell cultures and data from the transferrin endocytosis assays were analyzed by t test in Graphpad Prism; (ii) assessment of fold-change in FcgR mRNA expression post-exposure to IFNγ and of change in FcgR transcript expression in vivo were performed by built-in randomization algorithms in REST 2009 (separately for each litter), then fold-changes and 95% confidence intervals from each litter were copied into Graphpad Prism to calculate p values; and (iii) densitometric analysis of constitutive FcγR expression in cortical cells was performed in Graphpad Prism using two-way ANOVA.

We first examined the transcript profile of FcγR in the developing rat brain. FcgrIa, FcgrIIa, FcgrIIb, FcgrIIIa, and Fcgrt mRNA was quantified in the cortex, hippocampus, and cerebellum of P1 male and female pups using qPCR. The spleen and liver from these pups were used as control tissues. All five Fcgr transcripts were detected in all three brain regions and control tissues in both sexes (Fig.  1a ). Significant differences in transcript levels between males and females were not detected by 2-way ANOVA (testing for sex and tissue effects; p set at 0.05) in any of the tissues examined; therefore, data from both sexes were pooled for this and all subsequent analyses. Significant differences were observed between different FcgR transcript levels within each of the different tissues (F(1,4)=116.880, p = 0.000 for spleen, F(1,4)= 103.749, p = 0.000 for liver, F(1,4) = 54.082, p = 0.000 for cortex, F(1,4)=57.298, p = 0.000 for hippocampus, F(1,4)= 74.946, p = 0.000 for cerebellum), and Fcgrt was the most abundant transcript across all tissues and brain regions tested (p = 0.000; block design ANOVA with Tukey’s post hoc test; n = 6 pups from each of 2 independent litters). Transcript levels for each FcgR differed between tissues (F(1,4) = 52.436, p = 0.000 for FcgRIa, F(1,4) = 132.459, p = 0.000 for FcgRIIa, F(1,4) = 161.012, p = 0.000 for FcgRIIb, F(1,4) = 72.733, p = 0.000 for FcgRIIIa, F(1,4) = 254.119, p = 0.000 for FcgRt; block design ANOVA). All Fcgr transcripts were expressed at lower levels in the cortex, hippocampus, and cerebellum compared to the spleen (p = 0.000; Tukey’s HSD post hoc test). Fcgr levels also differed between brain regions. Specifically, the cortex expressed less FcgrIIa (p = 0.006) and FcgrIIb (p = 0.000) compared to the hippocampus, and cortical levels of Fcgrt were lower than in the hippocampus (p = 0.000) or cerebellum (p = 0.001). FcgrIIb was less abundant in the cerebellum than in the hippocampus (p = 0.000). No differences in FcgrIa or FcgrIIIa expression were observed between brain regions.

To assess whether Fcgr transcript levels changed as a function of age, Fcgr transcript levels in tissues from P7 pups (n = 11 pups from two litters, except for the spleen where n = 8 from two litters) were compared to levels quantified in P1 tissues (n = 12 pups from two litters). Fcgr transcript levels were generally increased in the brain of P7 pups relative to their P1 littermates (Additional file  1 ). Significant changes were observed in the cerebellum where FcgrIIb (2.661 fold; 95% CI, 0.513–23.645; p = 0.004) and Fcgrt (2.294 fold; 95% CI, 0.679–9.559; p = 0.001) increased from P1 to P7, and in the cortex, where FcgrIIb (2.491 fold; 95% CI, 0.343–22.422; p = 0.034) increased from P1 to P7. Larger confidence intervals for FcgrIIb and Fcgrt in the cerebellum are due to the variability in the Ct values obtained between different animals and could be due to regional differences in FcγR expression between different areas of the cerebellum (although the entire cerebellum was removed and stored in RNALater across all animals, the pieces that were used for analysis were randomly chosen). In contrast to the cerebellum and the cortex, in the spleen, FcgrIa (0.276 fold; 95% CI, 0.034–8.977; p = 0.018), FcgrIIa (0.244 fold; 95% CI, 0.077–3.074; p = 0.001), and Fcgrt (0.452 fold; 95% CI, 0.130–3.412; p = 0.009) decreased significantly from P1 to P7. Fcgr transcript levels did not change significantly between P1 and P7 in the hippocampus or liver. Results for liver and spleen can be found in Additional file  2 .

To determine whether FcgrIa, FcgrIIa, FcgrIIb, FcgrIIIa, and Fcgrt transcripts are also expressed in the developing rat brain in vitro, and particularly in non-immune cells, mRNA levels of these Fcgr genes were quantified by qPCR in primary hippocampal and cortical cells dissociated from P1 rats (pooled male and female brain regions). These cultures were comprised of MAP2b-positive neurons and GFAP-positive astrocytes and were confirmed to be devoid of microglia and endothelial cells, as determined by lack of immunolabeling for CD68 and Iba-1 (microglial markers) and CD31 (endothelial cell marker), and to contain < 1% CD11b immunopositive cells (Additional files  3  and  4 ). Consistent with in vivo expression patterns, on DIV 2, all five Fcgr transcripts were detected in hippocampal (Fig.  1b ) and cortical cell cultures (Fig.  1c ). There were significant differences in expression levels between different Fcgr transcripts in hippocampal cell cultures (F(1,4) = 12.379, p = 0.016, block design ANOVA; n = 3 cultures per group in each of two dissections). As observed in vivo, Fcgrt was expressed at significantly higher levels than either FcgrIa (0.060; p = 0.034) or FcgrIIIa (0.079; p = 0.013) (Fig. 1b). Small differences between transcript levels were also observed in DIV 2 cortical cell cultures (Fig.  1c; F(1,4) = 14.087, p = 0.013, block design ANOVA), with Fcgrt and FcgrIIa being more abundantly expressed than FcgrIIb (mean difference: 0.070; p = 0.023 and 0.066; p = 0.028, respectively) or FcgrIIIa (0.065; p = 0.029 and 0.061; p = 0.036, respectively). Comparison of each Fcgr transcript between the two culture types (t test) revealed a significant difference only for Fcgrt, which was slightly more abundant in hippocampal cell cultures compared to cortical cell cultures (mean difference = 0.025; p = 0.026; df = 2).

To further assess the ontogeny of Fcgr mRNA expression in vitro, Fcgr transcript levels were determined from total RNA isolated from primary hippocampal and cortical cell cultures on DIV 7, 14, and 21 and compared to levels on DIV 2. Significant changes in Fcgr expression as a function of DIV were observed in both culture types. The relative expression values along with 95% confidence intervals and p values for each gene in hippocampal and cortical cell cultures are listed in Table  2 (pooled data from 6 cultures derived from 2 independent dissections; REST2009 software built-in statistical analysis). In hippocampal cell cultures, FcgrIa, FgcrIIa, FcgrIIb, and FcgrIIIa transcript levels were decreased on DIV 7, 14, and 21 compared to DIV 2, while Fcgrt expression remained constant across all DIV. In cortical cell cultures, FcgrIa, FcgrIIa, FcgrIIb, and FcgrIIIa levels were lower on DIV 7 compared to DIV 2, while only FcgrIIa and FcgrIIIa levels were reduced on DIV 14 and DIV 21 compared to DIV 2. Fcgrt levels were significantly increased on DIV 14 but not changed on DIV7 or DIV21 compared to DIV 2.

Having confirmed that non-immune cells in the developing rat brain express Fcgr and that the transcript profile of Fcgr in vitro resembles that in vivo, we next determined whether FcγR are expressed at the protein level in primary hippocampal and cortical cell cultures. We first used western blotting to interrogate FcγR protein expression in rat cortical cell cultures. Commercially available antibodies specific for FcγRIa (CD64; protein product of FcgrIa), FcγRIIb (CD32; protein product of FcgrIIb), FcγRIIIa (CD16; protein product of FcgrIIIa), and FcRn (protein product of Fcgrt) were initially tested against cell lysates of adult rat peritoneal macrophages as a positive control. Bands immunoreactive for FcγRIa and FcγRIIIa of the anticipated size (40–45 kDa) were detected in peritoneal macrophage lysates (not shown). In contrast, anti-FcγRIIb and anti-FcRn antibodies failed to react with macrophage cell lysates, suggesting that these antibodies are not suitable for western blotting. Similarly, FcγRIa (Fig. 2a) and FcγRIIIa (Fig.  2b ), but not FcγRIIb or FcRn, were detected in cortical cell lysates collected on DIV 2, 7, 14, and 21. Densitometric analysis revealed no significant differences in FcγRIa or FcγRIIIa protein levels in cortical cells on DIV 7, DIV 14, and DIV 21 compared to DIV 2 (Fig. 2c, d; F(1,3) = 1.371, p = 0.4008 for FcγRIa; F(1,3) = 0.6472, p = 0.6353 for FcγRIIIa, 2-way ANOVA).

In immune cells, the responses triggered as a result of immune complex recognition by FcγR can vary from cell to cell depending on the ratio of excitatory to inhibitory FcγR expressed on each cell [34, 35]. We hypothesized that FcγR subtype expression at the protein level similarly differs between individual neurons and/or astrocytes. Using flow cytometry of cells dissociated from the neocortex of P0 rat pups (Fig.  3a ), we confirmed that cortical neurons, identified as MAP2b-immunopositive cells (Fig.  3b ), and astrocytes, identified as GFAP-immunopositive cells (Fig.  3c ), respectively, express FcγRIa (Fig. 3d, g) and FcγRIIb (Fig. 3e, h). Although some cortical neurons expressed either FcγRIa or FcγRIIb, a subset of cortical neurons expressed FcγRIa and FcγRIIb simultaneously at varying intensities (Fig.  3f ). Similarly, a subset of cortical astrocytes was immunopositive for both FcγRIa and FcγRIIb (Fig.  3i ).

To assess the subcellular expression of FcγR in neurons and astrocytes, we used immunocytochemistry to label hippocampal and cortical cells at DIV 7. Results confirmed our flow cytometry data, showing hippocampal and cortical neurons and astrocytes immunopositive for FcγRIa (Fig.  4 ) and for FcγRIIb (Fig.  5 ). Although MAP2b-immunopositive neurons were consistently immunoreactive for FcγRIa and/or FcγRIIb, immunolabeling of GFAP-immunopositive astrocytes for FcγR was highly variable between and within cultures. Some GFAP-immunopositive cells in our cultures did not react with either FcγRIa or FcγRIIb antibodies. Confocal microscopy confirmed that when cells were fixed and permeabilized before FcγR staining, they revealed intracellular (in some cases including intranuclear) staining for FcγRIa and FcγRIIb in addition to surface staining (not shown); this pattern of intracellular staining was not observed in cells which were immunolabeled before fixation. Intracellular immunolabeling for FcγR in fixed, permeabilized cells could be explained by the presence of intracellular pools of FcγR in neurons and astrocytes, available for translocation to the cell membrane or, conversely, reflect internalized receptors. The presence of neuronal surface receptors in intracellular pools [56, 57], as well as the presence of intracellular pools of FcγR in leukocytes [58], have been described before.

FcγRIIIa and FcRn immunoreactivities were not detected by immunocytochemistry in hippocampal or cortical cell cultures. Although cultures of adult rat peritoneal cells were immunoreactive for FcRn (data not shown), FcRn immunoreactivity was not observed in hippocampal and cortical cultures. The absence of FcγRIIIa immunostaining and the presence of FcγRIIb immunostaining in primary cortical and hippocampal neurons are in contrast to our western blot data, suggesting that the anti-FcγRIIIa and anti-FcγRIIb antibodies we used are not dual purpose antibodies that would be suitable for both western blotting and immunocytochemistry. This possibility was confirmed by observations that the anti-FcγRIIb antibody detected FcγRIIb immunoreactivity in peritoneal cells by immunocytochemistry but not by western blotting (data not shown).

To confirm surface expression of FcγR in cultures of neurons and astrocytes as detected by immunocytochemistry, we employed chemoproteomic cell surface capture (CSC) technology [53]. CSC technology enables surfaceome analysis and phenotyping of cells without antibodies by using a mass spectrometry-based strategy. Following selective cell surface labeling using a biotinylated probe, formerly N-glycosylated and cell surface-residing glycoproteins were identified. A non-proteotypic glycopeptide derived from either FcγRII or FcγRIII was identified in surfaceome preparations from hippocampal cell cultures at DIV 7, DIV 14, and DIV 21. Using the open source tool, Protter [59] (Fig.  6 ), the identified glycopeptide (ATVNDSGEYR) was shown to map to the extracellular domain of FcγRII (Uniprot ID Q63203; Fig.  6a ) and FcγRIII (Uniprot ID P27645; Fig.  6b ). These data provide additional and independent confirmation that the FcγR protein (FcγRII and/or FcγRIII) is located on the cell surface in hippocampal neuron-glia co-cultures.

IFNγ is an important inflammatory mediator produced by T cells and natural killer (NK) cells [60] that regulates FcγR expression in immune cells [61‐64]. IFNγ has been suggested to upregulate FcγR expression in adult rat microglia at the protein level [65], but whether IFNγ alters expression of Fcgr mRNA in other cell types of the rodent brain is not known. To address this question, DIV 2 rat cortical cell cultures were exposed to recombinant rat IFNγ at 100 ng/ml, a concentration previously shown to influence signaling in cultured cortical neurons without affecting cell viability [66]. As determined using the statistical analysis package in the REST2009 software (that includes built-in randomization algorithms) to assess qPCR data, a 24-h exposure of cortical cell cultures to IFNγ significantly increased levels of FcgrIa (4.271-fold; 95% CI 3.705–4.837), FcgrIIa (1.351-fold; 95% CI 1.008–1.695), and FcgrIIIa (6.525-fold; 95% CI 4.403–8.646) but decreased levels of FcgrIIb (0.326-fold; 95% CI 0.201–0.451) and Fcgrt (0.653-fold; 95% CI 0.455–0.850) relative to sister cultures treated with vehicle (0.1% DMSO) (Fig.  7a ). After 48 h of exposure to IFNγ (Fig.  7b ), expression of FcgrIa (4.297-fold; 95% CI 2.247–6.348), FcgrIIa (1.723-fold; 95% CI 1.397–2.049), and FcgrIIIa (5.624-fold; 95% CI 4.444–6.803) remained significantly elevated, and levels of Fcgrt (0.802-fold; 95% CI 0.614–0.990) were still significantly decreased. Grubb’s test identified one of the 6 independent experiments as an outlier for FcgrIIb transcript levels after a 48-h exposure to IFN; with this outlier removed, FcgRIIb transcript levels were confirmed to be significantly decreased, similarly to the 24 h exposure to IFNγ (0.326-fold; 95% CI 0.2279–0.4245).

In myeloid and lymphoid cells, cross-linking of excitatory FcγR, such as FcγRIa and FcγRIIIa, by IgG immune complexes (IgG-IC) triggers a rapid increase in intracellular calcium ([Ca²⁺]_i) levels [32]. In contrast, mobilization of intracellular calcium is not observed upon ligation of inhibitory FcγR, such as FcγRIIb [32]. It has been suggested that IgG-mediated activation of FcγR increases [Ca²⁺]_i in peripheral neurons [67, 68]. We adapted a previously published approach of bath application of IgG-IC to stimulate [Ca²⁺]_i in hippocampal cell cultures [54]. To obtain high-quality images of individual cells and to distinguish between IgG-mediated effects on [Ca²⁺]_i and spontaneous, synchronized calcium oscillations normally occurring in mature hippocampal neurons in culture [69, 70], calcium imaging studies were performed on low density DIV 7 hippocampal cell cultures pre-loaded with the calcium indicator dye Fluo4. IgG-IC consisting of pre-complexed mouse IgG and rat anti-mouse IgG was added to the culture medium at varying concentrations.

Calcium traces of individual cells showed that under the culture conditions used in these experiments, neurons did not exhibit synchronized calcium oscillations (Fig.  8a ). Analysis of calcium traces of individual cells indicated that exposure to IgG-IC caused concentration-dependent increases in [Ca²⁺]_i (Fig.  8b, c ). Although the shape of calcium traces in IgG-IC responsive cells varied between IgG-IC concentrations and between experiments for any experimental condition, including vehicle, they generally exhibited one of two general responses: (i) a spike in [Ca²⁺]_i or (ii) a gradual increase in [Ca²⁺]_i (Fig.  8a ). Quantification of the percentage of cells that responded to IgG-IC revealed that median values increased in a concentration-dependent manner up to 1 μg/ml (Fig.  8b ). Calculation of the average area under the curve (AUC) for calcium traces revealed a concentration-dependent increase up to 1 μg/ml IgG-IC (Fig.  8c ), whereas a reduction in average AUC was observed with 10 μg/ml IgG-IC and an increase was observed with 100 μg/ml IgG-IG. However, no statistically significant differences between means were found for either endpoint (block design ANOVA with statistical significance set at p = 0.05; n = 1–3 cultures in each of 4 independent dissections). The individual components of the IgG-IC were used in control experiments. Specifically, cells were incubated with either 100 μg/ml mouse IgG or with 100 μg/ml rat anti-mouse IgG and the percentage of responding cells as well as the average AUC of the calcium traces were calculated (Fig.  8b, c ). Both mouse IgG and rat anti-mouse IgG at a very high concentration of 100 μg/ml increased [Ca²⁺]_i in some cells (Fig.  8b ). The median AUC for calcium traces in response to 100 μg/ml mouse IgG was lower compared to 100 μg/ml IgG-IC, whereas the median AUC of the calcium traces at 100 μg/ml rat anti-mouse IgG was higher compared to 100 μg/ml IgG-IC (Fig.  8c ). However, these differences were not significantly different from untreated cells, as determined by block design ANOVA (statistical significance set at p = 0.050; n = 1–3 cultures in each of 3 independent dissections for mouse IgG alone and 2 independent dissections for rat anti-mouse IgG alone). Although we did not test the individual components of the IgG-IC at concentrations lower than 100 μg/ml, based on evidence that IgG alone (in the absence of immune complex) has been shown to act antagonistically on FcγR signaling in immune signaling, we would not expect to observe a robust calcium response at lower concentrations of the individual components of the immune complex [71, 72].

To exclude the possibility that these studies were confounded by varying percentage of astrocytes comprising the total cell population, cultures from 2 of the 4 independent dissections used for this study were immunostained for MAP2b and GFAP following completion of the calcium assay. Fluorescent images of the same fields in the cells that had been used for the calcium assay were acquired to compare calcium traces from neurons versus astrocytes. Analysis of calcium traces of neurons alone was conducted as above, and the results (not shown) were similar to those reported above, independent of cell type, e.g., the percentage of neurons responsive to IgG-IC was concentration dependent, but not significantly different from vehicle control.

In immune cells, increased [Ca²⁺]_i as a result of cross-linking of excitatory FcγR has been linked to activation of the MAPK signaling pathway and Erk phosphorylation [32]. Similarly, we observed that IgG-IC increased levels of phosphorylated Erk (pErk) in adult rat peritoneal cells in a concentration-dependent manner (Fig.  9a ). The ratio of pErk to Erk was increased significantly (60.44; p = 0.046; Tukey’s HSD) in DIV 7 hippocampal cell cultures exposed to IgG-IC at 10 μg/ml relative to control cultures treated with media alone but not at 100 μg/ml (Fig.  9b; (F(2,7) = 3.331, p = 0.026, block design ANOVA)). In contrast, when cells were exposed to rat anti-mouse IgG alone, 10 μg/ml had no effect on Erk phosphorylation whereas 100 μg/ml significantly increased the pErk/Erk ratio (70.05; p = 0.016; Tukey’s HSD) (Fig.  9b ). Exposure to IFNγ (30 ng/ml) had no significant effect on the pErk/Erk ratio in hippocampal cell cultures, and IFNγ did not enhance the effect of IgG-IC on this outcome after 30 min nor after 24 h of exposure (Fig.  9b and Additional file  5 ).

In macrophages, cross-linking of FcγR by immune complex often results in FcγR internalization into endosomes [73]. To determine whether this also occurs in neural cells, DIV 7 hippocampal cell cultures were “primed” using rat anti-mouse IgG conjugated to AlexaFluor 647, followed by addition of mouse IgG and AlexaFluor 488-labeled rat transferrin, which labels the entire endosome pathway [55]. Intracellular puncta dual-labeled for rat anti-mouse IgG and transferrin (Fig. 10a, c) were quantified as a measure of FcγR internalization. The number of puncta immunopositive for both rat anti-mouse IgG and transferrin was significantly increased in cultures exposed to the IgG-IC (p = 0.011; df = 5; mean difference = −0.7416 by t test of log transformed data), but not in cultures exposed to rat anti-mouse IgG in the absence of mouse IgG (Fig. 10b). This effect of IgG-IC was significantly reduced in hippocampal cell cultures pre-incubated with antibodies against FcγRIa (CD64) or FcγRIIb (CD32) (F = 5.92; p = 0.007; df = 28; Tukey’s HSD) (Fig. 10c, d).