Background
The response of the central nervous system (CNS) to disease, inflammation, and injury features prominent involvement of astrocytes, the "macroglia" population of the CNS [
1], as well as microglia, the primary resident immune cell population [
2]. While the astrocytic and microglial responses in these pathogenic contexts have been thought to involve cross-talk between these two cell populations [
3], the mechanisms and functional significances underlying these interactions are incompletely understood [
4,
5]. In the retina, Müller cells, the radial astroglia of the retina, as well as retinal microglia, similarly demonstrate marked changes in various retinal injuries and diseases [
6,
7]. Müller cells and microglial injury responses in the retina have been ascribed both beneficial and deleterious features [
8‐
10], and the "Janus-faced" nature of these responses may be secondary to the complex system of communications between microglia and macroglia. How these two cells types interact in the aftermath of retinal injury, and how they shape an adaptive or maladaptive overall response have not been fully explored.
In the uninjured state, microglia in the brain [
11,
12] and the retina [
13] demonstrate dynamic motility in their processes that are likely to subserve immune surveillance [
14]. Time-lapse imaging in live tissue has also shown that in the minutes following tissue injury, microglia in the vicinity of injury polarize their processes to the injury site and initiate positive chemotaxis and aggregation [
11‐
13]. Based on these and other observations [
15‐
17], the activation of microglia in the acute aftermath of CNS injury is generally thought to constitute the initial step of a generalized inflammatory response, and precede responses in astroglia. An understanding of how the overall injury response, initiated following microglial activation, is subsequently shaped by communications between these two cell types is likely central to elucidating the functional importance of post-injury inflammation and the mechanisms underlying chronic neuroinflammatory changes thought to drive CNS pathologies, including retinal disease [
10].
In the current study, we aimed to examine the existence, nature, and functional significance of microglia- Müller cell interactions following microglial activation. We employed an in vitro co-culture system where morphological, molecular, and functional alterations in cultured Müller cells were examined following juxtaposition with activated and non-activated microglial cells. We also sought corroboration for in vitro observations with in vivo examinations of microglia- Müller cell interactions in the retina following microglial activation. Our observations demonstrated that Müller cells exhibited prominent molecular and functional responses to microglial activation that appeared to be adaptive in nature and which differed from the changes found in chronic gliosis. We also found evidence for bidirectional signaling between microglia and Müller cells that facilitated further activation, migration, and cell adhesion of microglia that may be relevant to amplifying and coordinating an inflammatory response in the retina. These data underline a program of injury response in the retina constituted by microglia- Müller cell interactions that likely serve to provide protection to surrounding neurons and to restore tissue homeostasis.
Methods
Culture of mouse retinal Müller cells
Experiments were conducted according to protocols approved by the NEI Institutional Animal Care and Use Committee and adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Experimental animals were housed and bred in National Institutes of Health animal facilities.
Retinal cell cultures were obtained using the retinas of wild type C57BL/6 mice (Charles River Laboratories Inc., Wilmington, MA). Müller cells were isolated from the retinas of postnatal day (P)7-12 C57BL/6 mice using a protocol modified from Hicks and Courtois [
18]. Following euthanasia, mice were rapidly enucleated and their globes immersed in Dulbecco's modified Eagle's medium (DMEM) containing 1:1000 penicillin/streptomycin overnight in the dark at room temperature. These were subsequently transferred into 0.1% trypsin at room temperature for 60 minutes and then rinsed thrice with DMEM containing 10% fetal bovine serum (FBS, Gibco/Invitrogen, Carlsbad, CA) to terminate the digestion reaction. The retinas were carefully dissected free from the other ocular tissues and dissociated by trituration. The resulting cell suspension was then seeded into 75 cm
2 flasks (4-5 retinas per flask) containing DMEM medium with 10% FBS at 37°C. The culture medium (DMEM medium with 10% FBS) was changed 24 hours after seeding. At 3-4 days intervals, cultures were shaken vigorously to detach non-adherent cells which were then removed from the culture by aspiration. When the remaining adherent cells in the cultures reached 80% confluence, they were detached from the flask bottoms using 0.1% trypsin, resuspended in fresh DMEM containing 10% FBS, and replated into new flasks. Immunohistochemical staining of these remaining cells demonstrated them to be composed of > 98% Müller cells as evidenced by immunopositivity for glutamine synthetase (GS), glutamate aspartate transporter (GLAST), vimentin, as well as immunonegativity for CD11b (microglia), NeuN (neurons), and GFAP (astrocytes).
Co-culture of retinal microglia and Müller cells
Mouse retinal microglia cells were isolated from retina of young C57BL/6 mice (P10-30) and evaluated for purity as previously described [
19]. Cultured microglia were collected and seeded at a density of 0.5-1 × 10
6/well in 10% FBS DMEM onto Transwell permeable support membrane inserts (Corning, Corning, NY) and allowed to settle and grow for the next 24 hours. These cell-bearing inserts constituted the upper chamber in a two-chambered microglia-Müller cell co-culture system. Cultured Müller cells were seeded into the bottom of 6-well plates at a density of 0.5-1 × 10
5 cells/well in 10% FBS DMEM. After 24 hours, Müller cells were washed with DMEM and replaced with 0.5 ml of DMEM containing 5% heat-inactivated (HI) serum. These Müller cells were then co-cultured with microglia-containing inserts for 48 hours in the following configurations: (1) Müller cells incubated with empty inserts lacking microglia (control), (2) Müller cells incubated with inserts containing microglia that had not been exposed to LPS, and (3) Müller cells incubated with inserts containing activated microglia which had been previously exposed to 1 μg/ml of lipopolysaccharide (LPS) (1 μg/ml, Sigma, St Louis, MO) for 6 hrs and then rinsed thoroughly three times with DMEM to remove all residual LPS before co-culture.
Following co-culture, the two co-culture chambers were disassembled and exposed Müller cells were either: (1) washed with PBS and harvested for mRNA analysis, (2) fixed in 4% paraformaldehyde in PBS for 60 minutes for subsequent immunohistochemical and labeling studies, or (3) washed thrice with DMEM and incubated in fresh DMEM containing 5% HI-serum for an additional 24 hours to generate conditioned media.
Immunohistochemistry
Immunohistochemistry was performed on cultured Müller and microglial cells as well as mouse retinal sections. Cultured cells were fixed in 4% paraformaldehyde in PBS at room temperature for 30-60 minutes and washed thrice with PBS. Retinal sections were prepared from fixed eyecups of experimental animals which had been cryoprotected in 30% sucrose in PBS for 24 hrs, embedded in OCT (Tissue-Tek, Elkhart, IN), and flash frozen in acetone containing dry ice. Retinal sections of 30- μm thickness were cut on a cryostat (Leica CM 3050S, Buffalo Grove, IL) and stored at -20°C. Sections were thawed and rinsed with PBS prior to immunohistochemical staining.
Cultured cells or retinal sections were pre-incubated in blocking buffer (consisting of 10% normal goat serum (NGS), 5% bovine serum, and 0.5% Triton X-100 in 1 × PBS) for 2 hours at room temperature, and then in primary antibody (diluted in the blocking buffer) overnight at room temperature. Primary antibodies targeting the following molecules were used: glutamine synthetase (GS, clone: GS-6), 1:200; NeuN (clone: A60), 1:200; glutamate aspartate transporter (GLAST), 1:200 (all from Chemicon International, Temecula, CA); glial fibrillary acidic protein (GFAP, clone: 2.2b10), 1:200 for cell culture, 1:800 for retinal sections; CD11b (clone: 5C6), 1:200; F4/80 (clone: A3-1), 1:200 (all from AbD Serotec, Raleigh, NC); ionized calcium binding adaptor molecule-1 (Iba1), 1:800 (Wako, Richmond, VA); vimentin, 1:200 to 1:800 (clone: V9, MP Biomedicals, Solon, OH and AbCam, Cambridge, MA). The cells or sections were then washed 3 times with 1 × PBS, and incubated in secondary antibody (1:400) for 1 hr at room temperature. The following secondary antibodies were used: Alexa 488-conjugated goat anti-rabbit IgG antibody, Alexa 488-conjugated goat anti-chicken IgG antibody, Cy3- conjugated goat anti-mouse IgG antibody, Cy5- conjugated goat anti-mouse IgG, Cy3- conjugated goat anti-rat IgG and Cy3- conjugated goat anti-guinea pig IgG antibody (all from Invitrogen, Carlsbad, CA). F-actin in cultured cells was labeled using AlexaFluor 555-conjugated Phalloidin (1:100, from Invitrogen, Carlsbad, CA) by incubating for 2 hours at room temperature. After immunolabelling, cells or sections were rinsed thrice in 1 × PBS and coverslipped in Vectashield mounting medium containing 4',6-diamidino-2-phenylindole (DAPI) (Vector Lab Inc, Burlingame, CA).
Confocal microscopy and image analysis of microglia and Müller cells in retinal sections
Retinal sections following immunohistochemical staining were imaged with confocal microscopy (FluoView 1000, Olympus, Japan). Multiplane z-series were collected using a 40× or 63×, oil-immersion objective. Each z-series spanned 30 μm in depth, and comprised of 30-50 images per series, each spaced 0.6-1 μm apart. Image stacks were analyzed with an image analysis software package (Volocity, Perkin Elmer, Waltham, MA) to generate a 3-dimensional rendering of the in vivo spatial relationships between microglia and Müller cells.
Morphological analysis of cultured Müller cells
Cultured Müller cells following co-culture were immunolabeled for glutamine synthetase, phallodin and DAPI, and imaged with an epifluorescence microscope (Olympus BX51, Center Valley, PA). Computer-based analysis of the morphology of individual cells was performed with NIH ImageJ (Bethesda, MD, Version 1.441) using the "Analyze Particle" function. Morphological parameters included: cellular area, perimeter, circularity (defined as 4π(area of cell)/(cell perimeter2)), and elongation factor (defined as the ratio of the major axis to the minor axis of the cell).
Semi-quantitative rt-PCR
Lysates of cultured cells were homogenized by using QIAshredder spin column (Qiagen, Valencia, CA) and total RNA isolated using the RNeasy Mini kit (Qiagen) according to the manufacturer's specifications. Isolated RNA was treated with RNase-free DNase I (Qiagen) to remove any contaminating genomic DNA. First-strand cDNA synthesis from mRNA was performed using a cDNA synthesis kit (Ambion, Austin, TX) using oligo-dT as the primer. RT-PCR was performed using a 15 μl reaction cocktail in conjunction with 1 μl of the cDNA, 2 μl of primer mixture, and 12 μl of HotStarTaq Plus DNA Polymerase (Qiagen). After an initial denaturation at 95°C for 5 minutes, PCR amplification was conducted for 15-40 cycles. The cycle numbers for quantification of each product were chosen within the linear range of the PCR. GAPDH was used as an internal control. The stability of levels of GAPDH mRNA across experimental conditions was verified by comparison to two other house-keeping genes, β-actin and 18S rRNA. The absence of contaminating genomic DNA in the isolated RNA following DNase treatment was confirmed by performing the PCR amplification in the absence of reverse transcriptase in parallel control reactions and by ascertaining that no amplification product was formed. The primer sequences for the PCR reactions are provided in Table
1.
Table 1
DNA primers used in rt-PCR amplification reactions
bFGF | atggctgccagcggcatcac | tcagctcttagcagacattgga |
BDNF | ggactctggagagcgtgaa | ggtcagtgtacatacacagg |
CX3CL1 | ttcacgttcggtctggtggg | ggttcctagtggagctaggg |
CCL2 | gcaggtccctgtcatgctt | ctagttcactgtcacactgg |
CCL3 | ccaagtcttctcagcgccat | ggttgaggaacgtgtcctg |
CNTF | ggctttcgcagagcaatcac | gcttggccccataatggct |
GAPDH | cctctggaaagctgtggcg | gttgctgtagccgtattcatt |
GDNF | atgggattcgggccacttgg | tcagatacatccacaccgtttag |
GFAP | ttcctgtacagactttctcc | cccttcaggactgccttagt |
Glast | ctctgggcatcctcttcttg | caaatctggtgatgcgtttg |
GS | tgcctgcccagtgggaatt | tattggaagggttcgtcgcc |
ICAM-1 | ccaattcacactgaatgccag | ggcttgtcccttgagttttatg |
IFN-γ | ctggtggttgctcctcttac | ctcctgggcctctcctgtg |
IL-1β | gccaccttttgacagtgatgag | ttaggaagacacagattccatg |
IL-6 | atgaagttcctctctgcaagag | ctaggtttgccgagtagatctc |
iNOS | cctcccagccttgcatcc | cagagcctcgtggctttgg |
LIF | aatgccacctgtgccatacg | caacttggtcttctctgtcccg |
NGF | ggcgtacaggcagaaccgta | cagcctcttcttgtagccttc |
TGF-β | ccactgatacgcctgagtg | gctgcacttgcaggagcg |
TNF-α | atgagcacagaaagcatgatc | tcacagagcaatgactccaaag |
VCAM-1 | ggataccagctcccaaaatcc | cactttggatttctgtgcctc |
Vimentin | gtacaagtccaagtttgctg | atcgtgatgctgagaagtct |
Proliferation and apoptosis assays
Proliferating Müller cells and microglia were marked by the incorporation of bromodeoxyuridine (BrdU). BrdU was added to the culture medium for 2 hours (final concentration 10 μM), after which the cells were fixed in 4% paraformaldehyde for 60 minutes, washed with PBS, and incubated in 2N HCL in 1 × PBS for 60 minutes. After rinsing in 1 × PBS, cells were labeled with an anti-BrdU antibody overnight (G3G4, 1:200, Developmental Studies Hybridoma Bank, Iowa City, IA), followed by a goat anti-mouse antibody conjugated with Cy3 (Invitrogen, Carlsbad, CA). Cells were coverslipped in Vectashield mounting medium containing DAPI and imaged by epifluorescence microscopy. Images of 5 randomly chosen high-power fields were obtained for each well; the numbers of BrdU+ and DAPI+ cells in each field were counted and the percentage of the BrdU+ of the total cells present were calculated and averaged.
Cellular apoptosis in cultures was assessed by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) using an assay kit according to the manufacture's instruction (Roche, Nutley, NH). A similar mounting and imaging procedure was performed as for BrdU labeling and the percentages of TUNEL-positive cells among total cells present were calculated and averaged.
Photoreceptor neuroprotection assay
661W cells (a gift from Drs Zhongshu Tang and Xuri Li, National Eye Institute, Bethesda, MD), a photoreceptor-derived cell line, were plated in 96-well plates at a density of 1 × 104 cells/well. Hydrogen peroxide (H2O2, 0.1 mM) was added to induce oxidative stress-mediated photoreceptor cell death. The neuroprotective effects of Müller cell-conditioned media, as well as exogenous GDNF (AbCam, Cambridge, MA) and LIF (Prospec, East Brunswick, NJ) were assessed by adding them to H2O2 -stressed 661W cell cultures. Photoreceptor cell survival following 24-hour incubation was assessed using two separate methods according to the manufacturers' protocol: (1) a colorimetric microplate, tetrazolium salt-based, cell survival assay (Cell Counting Kit-8 Assay, Dojindo Molecular Technologies, Rockville, MD), and (2) a fluorescence-based, differential staining technique of live versus dead cells (LIVE/DEAD® Fixable Dead Cell Stain Kit, Invitrogen, Carlsbad, CA). The percentageof survival cells were quantified and expressed as a percentage of the untreated control (not subject to H2O2 -mediated stress).
ELISA
The protein levels of GDNF, LIF and CCL3 were measured using the following ELISA kits respectively according to their manufactures' instructions: 1) GDNF (Promega, Madison, WI), 2) LIF (R&D System, Minneaplos, MN), and 3) CCL3 (AdCam, ab100726, Cambridge, MA). Levels of IL-1β, IL6, TGF-β, ICAM and VCAM were measured with a commercial chemiluminescence-based ELISA assay (Searchlight, Aushon Biosystems, Billerica, MA).
Nitrite concentration assay
NO was measured indirectly via its stable metabolite nitrite using the Griess reagent system (Promega, Madison, WI). The conditioned media was first incubated with sulfanilamide solution for 10 minutes at room temperature. N-naphthylethylenediamine was added to the mixture and incubated for 10 minutes. The absorbance was read in a plate reader at 550 nM and nitrite concentration was calculated by comparison to a standard reference curve.
Evaluation of the effects of Müller cell-conditioned media on microglia
Fresh cultured microglia were plated in 6-well plates at the density of 1 × 106/well. After 24 hours, the cells were washed 3 times with DMEM and incubated in the conditioned media collected from Müller cells following co-culture for another 24 hours. Microglia were harvested for mRNA analysis or fixed in 4% paraformaldehyde in PBS for 60 minutes for a proliferation assay.
Microglial-Müller cell adhesion assay
The adhesion of microglia to Müller cells were assessed using an in vitro cell adhesion assay according to the manufacturer's instructions (Vybrant Cell Adhesion Assay Kit, Invitrogen, Carlsbad, CA). Müller cells were cultured similarly as before as: (1) alone, (2) co-cultured with unactivated microglia, or (3) co-cultured with activated microglia in 96-well transwell plates (Corning, Corning, NJ) with a seeding density of 5-8 × 103 cells/well in the lower chamber and microglia at a density of 0.5-1 × 104cells/well in the upper chamber for 48 hours. The upper chamber was then removed leaving behind the post- cocultured Müller cells in the bottom chamber. Separately, fresh unactivated cultured microglia were collected, washed with PBS, resuspended in DMEM containing 5 μM Calcein-AM at the density 1-2 × 106 cells/ml for 30-60 minutes to allow for cell labeling. The suspension of labeled microglia was centrifuged, washed with PBS, and then resuspended at the density of 2 × 105 cells/ml in DMEM. A volume of 150 μl of the Calcein-AM-labeled cell suspension was added to the previously co-cultured Müller cells in the well-bottoms and incubated at 37°C for 90 minutes to allow for microglia-to- Müller cell adhesion. The wells were carefully washed 4 times with DMEM to remove non-adherent cells and 200 μl of PBS was added. Fluorescence in each well was measured using Spectramax MT (Molecular Devices Inc., Sunnyvale, CA) and expressed as a fraction of the fluorescence of the total cells added to each well.
Microglial chemotaxis assay
Microglial chemotaxis was evaluated using 96-well transwell plates (Corning Incorporated, Corning, NY). The bottoms of the upper well-inserts contained polyester filters (8- μm pore size) that allowed for migration of cells from the upper chamber/well into the bottom chamber. Cultured microglia were harvested from cultures using 0.25% Trypsin-EDTA-mediated dissociation, centrifuged and washed with DMEM 3 times, resuspended in a 75 μl-volume of DMEM with 5% HI-FBS at a concentration of 1-2 × 105 cells/ml, and then seeded into the upper well insert. Conditioned media (125 μl) from previously co-cultured Müller cells were transferred into the bottom chamber. Additional bottom chambers also contained (1) non-conditioned medium (DMEM without serum) as a negative control, and (2) CCL2 (100 ng/ml in PBS) as a positive control. After incubation for 2 hrs at 37°C to allow for microglial migration, the top side of the insert filter was wiped carefully with a Q-tip to remove superficially adherent cells and the filter was removed. The filter was fixed in 4% paraformaldehyde, washed with PBS, stained with DAPI to mark the cells in the filter, and then mounted on a slide. Labeled cells that have migrated to the bottom side of the filter were imaged and counted.
Intravitreous injections in experimental animals
Adult C57BL/6 mice were anesthetized with an intraperitoneal injection of ketamine (90 mg/kg) and xylazine (8 mg/kg). A small penetrating incision into the vitreous cavity was made in the eye directly behind the limbus using a 30 gauge needle under a dissecting microscope (Olympus, model SZX16; Central Valley, PA). A 1 μl volume of 1 × PBS containing different concentrations of lipopolysaccharide (LPS) was injected the vitreous using a Hamilton microliter syringe (Hamilton Company, model 701, Reno, NV). Animals were euthanized 3 days after intravitreal injection and ocular tissues obtained for analysis.
Statistical analysis
Statistical analyses were performed using statistical software (Graphpad, San Diego, CA, USA). Comparisons between 3 or more data groups were performed using one-way analyses of variance (ANOVA), and comparisons between pairs of group means performed with the Tukey-Kramer multiple comparison test. A P value < 0.05 was set as the basis for rejecting the null hypothesis (i.e. that the group means for data groups being compared do not differ significantly from each other). In all graphical representations, the error bars indicate standard error (SE).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MW was involved in study design, performed the experiments, analyzed the data, images, and statistics, and wrote the manuscript. WM contributed to cell-culture experiments and helped write the paper. LZ performed the in vivo experiments and helped write the manuscript. RNF contributed to imaging experiments and data analysis, and helped write the manuscript. WW was involved in study design, provided research support, and helped write the manuscript. All authors have read and approved the final version of the manuscript.