- Email: [email protected]
Purification of astrocytes from adult human optic nerve heads by immunopanning
Brain Research Protocols 12 (2003) 67 – 76 www.elsevier.com/locate/brainresprot
Protocols
Purification of astrocytes from adult human optic nerve heads by immunopanning Ping Yang a, M. Rosario Hernandez a,b,* a
Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, St. Louis, MO, USA b Department of Anatomy and Neurobiology, Washington University School of Medicine, St. Louis, MO, USA Accepted 16 July 2003
Abstract Most in vitro studies in the CNS require pure cultures of astrocytes. Astrocytes from the human optic nerve head (ONH, type 1B) represent a specialized population of astrocytes. Primary cells grown from human optic nerve head explants were cultured for 3 – 4 weeks. To select astrocytes by immunopanning, cell suspensions were placed on a P100 panning dish coated with C5 anti-neuroepithelial antibody and allowed to attach for 30 min. Nonadherent cells were plated on a second dish coated with anti-Thy1.1 antibody to deplete microglia and meningeal cells. Finally, remnant nonadherent cells were plated on a noncoated dish. Purified cells were immunostained with astrocyte markers: GFAP, vimentin, Pax2, A2B5, nestin and NCAM. Other cell types were characterized by HLA-DR for microglia and smooth muscle actin for vascular smooth muscle. The proportion of GFAP+ astrocytes in the cultures was determined by flow cytometry. About 95% of the cells that adhered to the C5 dish were GFAP+ astrocytes. GFAP+ astrocytes expressed vimentin, Pax2, nestin and NCAM, but not A2B5. From the Thy1.1 dish, 60 – 75% cells were GFAP+ astrocytes and the remainder cells were GFAP cells. Using cloning rings, we eliminated fibroblast-like cells, smooth muscle and meningeal cells from astrocyte cultures. Smooth muscle cells and fibroblasts grew on the noncoated dish. In conclusion, immunopanning is an efficient method to get high yields of viable type 1B astrocytes from adult human ONH. The current described culture system may provide a valuable tool in studying human optic nerve head biology and disease. D 2003 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Astrocyte, optic nerve head Keywords: Human optic nerve head; Astrocyte; Explants culture
1. Type of research
Astrocytes are the major glial cell type in the nonmyelinated optic nerve head (ONH) in most mammals and provide cellular support functions to the axons of the retinal ganglion cells while interfacing between connective tissue surfaces and surrounding blood vessels. In the ONH, transition from quiescent to reactive astrocyte occurs in response to various forms of stress and in disease such as proinflammatory cytokines, injury, and in degenerative diseases such as glaucomatous optic neuropathy.
To study cellular responses of ONH astrocytes in physiologic or pathophysiological conditions, we describe here a culture system to obtain purified OHN astrocytes (type 1B) from human postmortem eyes by immunopanning. Characterization of astrocyte cultures by immunocytochemistry and flow cytometry.
2. Time required
* Corresponding author. Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, 660 S. Euclid Avenue, Campus Box 8096, St. Louis, MO 63110, USA. Tel.: +1-314-7471448; fax: +1-314-747-1405. E-mail address: [email protected] (M.R. Hernandez). 1385-299X/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S1385-299X(03)00073-4
Preparation of reagents, 1.5 h Dissecting ONH and explant preparation: 2 h Explant culture: 1 –3 weeks Preparation of immunopanning dishes: overnight Dissociation of cells from explant culture: 40 min Panning procedure: 1.5 h Culture of cells in panning dishes: 1 –2 weeks Trypsinization and expansion of cultures: 1 –2 weeks
68
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Immunocytochemistry: 4 h Flow cytometry: 4 h Freezing stock of cells for further use: 2 h.
3. Materials
3.1. Human eye tissue
Human eyes without history of eye disease, neurodegenerative disease or diabetes (age 15 – 46) were obtained within 24 h of death from Mid-America Transplant Services (St. Louis, MO) and National Disease Research Interchange (NDRI, Philadelphia, PA). No donor consent was required for eyes used in research.
3.2. Special equipment
Culture dishes (100 and 35 mm; Falcon, Lincoln Park, NJ) six-well plates (Falcon) Primaria Tissue culture flasks (25 cm2, Falcon) Primaria Tissue culture flasks (75 cm2, Falcon) Pipettes (10, 5, 2 ml, Fisher Scientific, Pittsburgh, PA) Cover glasses (22 mm, Fisher Scientific) Stericup filter (Millipore, Fisher Scientific) Centrifuge tubes (50 ml, Fisher Scientific) Cell strainer (70 Am, Becton Dickinson, Franklin Lakes, NJ) Forceps (large, small) (Storz, St. Louis, MO) Surgical scissors (Storz) Surgical blades, 15 mm (Feather Safety Razor, Japan) Dissection microscope (Carl Zeiss, Thornwood, NY) Phase-contrast reversal microscope (Olympus CK2, Japan) TMAX black and white film (Kodak, Rochester, NY) Nikon Optiphot-2 microscope (Tokyo, Japan) Digital camera (Spot Diagnostic Instruments, Sterling Heights, MI) FACScan flow cytometer/CELLQuest Software system (Becton-Dickinson, San Jose, CA).
3.3. Chemicals and reagents
Hank’s balanced salt solution (HBSS, Washington University Medical School Tissue Culture Support Center, St. Louis, MO) Dispase (Sigma, St. Louis, MO) Dulbecco’s modified Eagle’s medium (DMEM)/F-12 (Washington University Medical School Tissue Culture Support Center) Fetal Bovine Serum (FBS, Biowhittaker, Walkerswille, MD) PSFM (10,000 U/ml penicillin, 10,000 Ag/ml streptomycin and 25 Ag/ml amphotericin B) (Gibco/BRL, Gaithersburg, MD)
1.3% bovine albumin fraction V (Sigma) ITS+ Premixk culture supplement (Collaborative Biomedical Products, Bedford, MA) Affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml, Jackson ImmunoResearch Lab, West Grove, PA) C5 neuroepithelial monoclonal antibody (1:20, gift from Dr. Ben Barres, Stanford University) Thy1.1 monoclonal antibody (1:1000, Chemicon, Temecula, CA) Cell suspension solution (Sigma) Rabbit anti-human GFAP (1:100, Sigma) Goat FITC-conjugated anti-rb IgG (1:200, Sigma) Donkey serum (Sigma) Vectashield (Vector, Burlingame, CA) PBS (Invitrogen) BSA (Fisher) Tris (50 mM, pH 9.5, Fisher) Alexa Fluork goat anti-rabbit or anti-mouse IgG (H + L) 488 (1:600, Molecular Probes, Eugene, OR) Alexa Fluork goat anti-rabbit or anti-mouse IgG (H + L) 568 (1:800, Molecular Probes) Texas Red-conjugated anti-mouse IgM (1:100, Chemicon) FITC-conjugated anti-mouse IgM (1:50, Chemicon).
4. Detailed procedures 4.1. Dissection Ten normal human eyes without history of eye disease, diabetes or chronic neurodegenerative disease (age 15– 46) were obtained within 24 h of death from eye banks throughout the United States through National Disease Research Interchange (NDRI) and the Mid-America Transplant Services. Eyes were transported to the laboratory in special sterile containers in a humidified atmosphere on ice, placed in sterile 100-mm Petri dishes and carefully rinsed with sterile cold 10-ml Hank’s balanced salt solution. The optic nerve head was freed from sclera and other neighboring tissues using surgical scissors and forceps and then transferred to a 35-mm Petri dish with 2 ml HBSS/10% Dispase. Under a dissecting microscope, pigmented and remnant nonneural tissues were removed and the post laminar myelinated nerve was identified and removed using a sharp blade. The resulting cylinder of tissue was bisected and the central vessels removed. The tissues were then cut into four explants. 4.2. Explants culture Tissue culture flasks (25 cm2) were conditioned with 0.5 ml growth medium: Dulbecco’s modified Eagle’s medium (DMEM)/F-12 supplemented with 10% FBS and PSFM (10,000 U/ml penicillin, 10,000 Ag/ml streptomycin and 25 Ag/ml amphotericin B), and prewarmed in a 37 jC, 5% CO2 incubator. Each explant was rinsed once in growth
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
medium, placed into the culture flask, and kept standing in the hood for 10 min until the explant adhered to the surface of the flask. Before placing in the incubator, 0.5 ml DMEM/ F-12/10% FBS was added to each flask. The growth medium was changed twice a week. When the first cells began to grow out of the explant, the medium was increased to 2 ml. One to three weeks later, the explants were removed and the cultures were fed with astrocyte-defined serum-free medium containing DMEM, 1.3% bovine albumin fraction V, 1 Al/ml ITS+ Premixk culture supplement and PSFM. Cells were allowed to grow to cover an area around the explants. Approximately, 10,000 cells are obtained per explant. Cells from three or four explants (30 – 50,000 cells) from the same eye were combined for panning. 4.3. Immunopanning Immunopanning was performed about one week after removing the explants. The immunopanning dishes were prepared the day before panning as described in detail by Mi and Barres [12]. Two 100-mm Petri dishes were incubated overnight at 4 jC in 10 ml Tris (50 mM, pH 9.5) containing affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml). The next day, each dish was washed three times with PBS. Following wash, 10 ml primary antibody in PBS/BSA (0.2%) was added and incubated at RT for 1.5 h. One dish was incubated with C5 neuroepithelial monoclonal antibody (1:20, gift from Dr. Ben Barres, Stanford University) and the second dish with Thy1.1 monoclonal antibody (1:1000). Afterwards, the dishes were rinsed with PBS, three times. Finally, 10 ml PBS was added to coat the dishes. Remove the PBS just before use. A cell suspension was prepared after removing media from T25 flasks. Briefly, 4 ml cell suspension solution was added to each flask and placed in the incubator for 40 min, then detach the cells by pipetting them up and down. The cell suspension was transferred to a 50-ml centrifuge tube and the cells spun down at 1100 rpm. The cells were then resuspended in 10 ml DMEM/F-12/10% FBS and filtered through a 70-Am pore cell strainer. A 200-Al aliquot was used to count cells and the rest of the cell suspension was incubated on the C5-coated dish for 30 min at room temperature in the hood with gentle agitation at 15 min to ensure access of all cells to the panning surface area. Next, the nonadherent cells were transferred to the Thy1.1-coated dish for 30 min. The C5 dish was gently washed with DMEM/HEPES for three to five times to remove the nonadherent cells. This step was monitored under the inverted microscope to make sure cells do not dry out and only the nonadherent cells were removed. The times of wash are important. Too many washes will wash off astrocytes, too few washes will not wash off the nonadherent cells completely. Afterwards, 10 ml DMEM/F-12/10% FBS medium was added to the C5 dish and placed in the incubator. The procedure with the Thy1.1 dish was the same. After panning, cells were
69
allowed to reach confluence in 1 – 2 weeks. Clusters of cells that were obvious contaminants were removed at this point using trypsinization (0.05% trypsin for 10 min) and cloning rings under the microscope. The purified cultures were then expanded by passing into T75 flasks. Samples of each culture were plated on coverslips for characterization by immunocytochemistry. The morphology of live cells was observed using an Olympus CK2 phase-contrast microscope. Images were recorded using photographic camera (Nikon NFX-35) using TMAX 400, black and white film. 4.4. Flow cytometry For flow cytometry, cells were grown in 75-cm2 flasks until confluence. Cells were then trypsinized and spun down at 4 jC. Cell pellets were washed three times by PBS. After removing PBS, the cells were fixed by adding 3 ml 4% paraformaldehyde for 30 min at room temperature. After two washes in PBS, cells were permeabilized in PBS containing 5% FBS, 0.2% Triton X-100 and 0.5% glycine for 30 min. The cells were divided into three 15-ml centrifuge tubes (f 2 106 cells per tube), one for sample (primary and secondary antibody present), one control for autofluorescence (without primary and secondary antibodies) and one for negative control (only secondary antibody). The sample was incubated with 150 Al primary antibody (rabbit anti-human GFAP 1:100) in PBS/1% BSA/0.5% Triton X-100 at room temperature for 45 min. After washing with PBS for three times, 150 Al secondary antibody, goat FITC-conjugated anti-rabbit IgG (1:200) in PBS/1% BSA was added to the sample and negative control. Cells were incubated in secondary antibody for
Table 1 Primary and secondary antibodies for immunocytochemistry Primary antibodies
1j Ab Source
1j Ab Dilution
Secondary antibodies
Mouse anti-human GFAP Mouse anti-human Vimentin Rabbit anti-human Nestin Mouse anti-human SMA Rabbit anti-human Pax2 Mouse anti-human NCAM Mouse anti-human A2B5 Mouse anti-human HLA-DR Mouse anti-human Thy1.1
Sigma
1:50
Sigma
1:50
Gift, Dr. Lundkvist Biogenex
1:1000 1:30
BabCO
1:50
Accurate
1:50
Chemicon
1:200
Accurate
1:50
Chemicon
1:200
Alexa goat anti-mouse IgG (H + L) 568 (1:800) Texas Red-conjugated anti-mouse IgM (1:100) Alexa goat anti-rabbit IgG (H + L) 488 (1:600) Alexa goat anti-mouse IgG (H + L) 568 (1:800) Alexa goat anti-rabbit IgG (H + L) 488 (1:600) Texas Red-conjugated anti-mouse IgM (1:100) FITC-conjugated anti-mouse IgM (1:50) Alexa goat anti-mouse IgG (H + L) 568 (1:800) Alexa goat anti-mouse IgG (H + L) 568 (1:800)
70
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
4.5. Immunocytochemistry For characterization, cells were cultured in 6 35-mm well plates containing 22-mm glass coverslips in growth medium until confluence. Cells were characterized by staining with astrocyte markers: glial fibrillary acidic protein (GFAP), vimentin, Pax2, nestin, neural cell adhesion molecule (NCAM) and A2B5. Other markers used included HLA-DR for microglia and smooth muscle actin (SMA) for vascular smooth muscle. Primary and secondary antibodies are summarized in Table 1. All primary antibodies were diluted in a blocking agent (10% donkey serum in PBS/ 0.5% BSA). All secondary antibodies were prepared in PBS/0.5% BSA.
Fig. 1. Phase-contrast micrographs of live primary cultured cells (age 15) from optic nerve head. (A) Cells grown on the C5 panning dishes have starshaped morphology. (B) Most of the cells grown on the Thy1.1 panning dishes have the same morphology as A, arrows point to large cells represent most likely lamina cribrosa cells. (C) Nonadherent cells were thin, elongated fibroblast-like cells or smooth muscle cells. Objective: A and B, 10 ; C, 20 .
45 min at room temperature. After washing, stained cells were maintained in the dark until analysis. The stained cells were measured using a FACScan flow cytometer/CELLQuest Software system. A single, aircooled, argon-ion laser provided excitation. Data were collected using logarithmic amplification on 5000 cells, excluding cell debris by a combination of forward and side scatters. Frequency histograms representative of cells obtained after panning with C5 and Thy1.1 are shown (Fig. 2). At least 10 cultures generated by panning from different donors have been tested.
Fig. 2. Flow cytometric analysis of cells adherent to C5 and Thy1.1 panning dishes. Percentage of GFAP-positive cells (after subtracting for nonspecific immunofluorescence, dotted line) adherent to C5 (A) and Thy1.1 (B) panning dishes.
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Cells on coverslips were washed with cold PBS once and fixed with cold 4% paraformaldehyde for 10 min, then washed three times by adding 1.5 ml of ice-cold PBS for 5 min. For cytoskeleton staining, cells were permeabilized in 0.1% Triton X-100 for 10 min. After blocking by incubation with 100 Al 10% donkey serum in PBS/ 0.5% BSA for 30 min, 100 Al of primary antibody was added to the coverslips and incubated at room temperature for 1 h. Cells were then washed three times in PBS and incubated with 100 Al appropriate secondary antibody for 45 min. After washing three times in PBS, the coverslips were mounted with Vectashield. For control studies, the primary antibody was replaced with nonimmune serum or omitted. Fluorescent images of immunolabeled cells were obtained using a Nikon Optiphot-2 microscope equipped with appropriate fluorescence filters for single or double fluorescence. Images were recorded using a digital camera and stored as computer files.
71
5. Results 5.1. Primary cell culture from human optic nerve head Initial outgrowth from the explanted lamina cribrosa appeared within 1 – 3 weeks in culture. Cells formed a monolayer around the explant and consisted of a mixture of flat polygonal cells with and without processes, and large flat cells with oval clear nuclei. Fibroblast-like cells formed independent clusters of elongated cells. 5.2. Immunopanning To purify astrocytes from explant culture, we used and modified the sequential panning procedure reported originally by Mi and Barres [12] and Barres et al. [2]. Briefly, the cell suspension of primary cells was panned first on a Petri dish coated with C5 anti-neuroepithelial monoclonal antibody to select for cells of astrocyte lineage. The nonadher-
Fig. 3. Immunofluoresence staining of cells (age 46) adherent to C5 panning dish. Positive intracellular staining of intermediate filaments of GFAP (A), vimentin (B) and nestin (C). Positive cell surface staining of NCAM (D) and Thy1.1 (E). Positive nuclear staining of Pax2 (G). Negative staining of A2B5 (F), SMA (H) and HLA-DR (I).
72
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
ent cells from the C5 dish were then passed onto a dish coated with the anti-Thy1.1 monoclonal antibody, a cell surface molecule member of the immunoglobulin superfamily [4] to deplete microglia, fibroblasts and meningeal cells. The nonadherent cells from the Thy1.1 dish were discarded. Almost all cells attached to the C5-coated dish were starshaped flat cells, with short processes. Very few other contaminant cell types were present (Fig 1A). Once the cells attached to the anti-C5-coated dish reached confluence, cells were detached by trypsinization and then plated on coverslips for characterization by immunocytochemistry and to be expanded for flow cytometry testing and further use in the laboratory. Interestingly, many cells attached to the anti-Thy1.1coated dish were also star-shaped cells, similar to the C5 dish (Fig 1B). Clusters of flat, small polygonal cells and isolated large cells with large nuclei represent most likely lamina cribrosa cells and meningeal cells, respectively. These cells were morphologically distinct from astrocytes
and they were removed from the culture dish using cloning rings and trypsinization (0.05% trypsin for 10 min). Once cells attached to the anti-Thy1.1 dish reached confluence, they were plated on coverslips for further characterization. The nonadherent cells from the anti-Thy1.1 dish were thin, elongated, fibroblast-like cells or smooth muscle cells (Fig 1C). These cells were discarded. 5.3. Flow cytometry In order to evaluate the purity of the astrocyte cultures derived from the anti-C5 and the anti-Thy1.1 dishes we performed flow cytometry. Astrocytes were cultured to confluence in astrocyte-defined serum-free medium for at least 10 days before flow cytometry. After labeling with GFAP and fluorescein, at least 95% cells derived from the anti-C5 dish were GFAP-positive astrocytes. Sixty to seventy-five percent of the cells derived from the anti-Thy1.1coated dish were also astrocytes (Fig 2).
Fig. 4. Immunofluoresence staining of cells (age 46) adherent to Thy1.1 panning dish. Positive intracellular staining of intermediate filaments of GFAP (A), vimentin (B) and nestin (C). Positive cell surface staining of NCAM (D) and Thy1.1 (E). Positive nuclear staining of Pax2 (G). Few cells have positive staining of A2B5 (F) and SMA (H). Negative staining of HLA-DR (I).
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
5.4. Immunocytochemistry The cytoskeleton of cells derived from the anti-C5coated dish stained for the intermediate filaments, GFAP, vimentin and nestin (Fig. 3). More than 95% cells derived from the anti-C5-coated dish were GFAP- and vimentinpositive. Few GFAP+ astrocytes (less than 10%) derived from the anti-C5 dish expressed nestin. Pax2, a transcription factor expressed by astrocyte lineage cells, labeled the nuclei of almost all cells derived from the anti-C5 dish. HLA-DR, a marker of microglia, was not expressed in cells derived from the anti-C5 dish. There were no cells that expressed smooth muscle actin, a marker of vascular smooth muscle. The cytoskeleton of 60 –75% cells derived from the antiThy1.1 dish stained with GFAP and vimentin (Fig. 4). Approximately 25 –35% of the cells expressed nestin and GFAP. All GFAP+ cells expressed Pax 2. Small clusters of cells expressed smooth muscle actin and they are presumably smooth muscle cells or pericytes. No microglial cell derived from the anti-Thy1.1 dish. NCAM is a cell surface adhesion molecule expressed by type 1B astrocytes of the human optic nerve head in vivo and in vitro [18,19]. Almost all GFAP+ cells derived from the C5 dish expressed NCAM as fine granules on the cell surface (Fig. 3). In addition, these cells reacted with the antiThy1.1 antibody in a similar pattern with anti-NCAM antibody. No cells derived from the anti-C5 dish expressed A2B5, a cell surface marker that labels type 2 astrocytes and precursors of the oligodendrocyte lineage. Between 50% and 75% of the GFAP+ cells derived from the Thy1.1 dish expressed NCAM and all cells expressed Thy1.1 (Fig 4). Few isolated GFAP+ cells derived from the Thy1.1 dish stained with the cell surface marker A2B5 (Fig. 4).
6. Discussion Here we describe a procedure to culture highly purified human optic nerve head astrocytes (type 1B) using a immunopanning method for rat optic nerve astrocytes modified from Mi and Barres [12] and Barres et al. [2]. Type 1B astrocytes are the major glial cell subpopulation in the nonmyelinated optic nerve head [24]. The ONH is thought to be the site of retinal ganglion cell (RGC) damage in primary open angle glaucoma (POAG), a common blinding neurodegenerative disease [22]. In glaucoma, there is extensive remodeling of the extracellular matrix (ECM) of the ONH with increased expression of ECM and proteolytic enzymes, cell adhesion molecules and neurotoxic mediators by reactive ONH astrocytes [1,11,16,21]. Therefore, ONH astrocytes represent a possible target for neuroprotection in glaucomatous optic neuropathy [10,17,23].Our laboratory developed techniques to culture type 1B astrocytes, but the yield, purity and viability of the cultures after multiple
73
passages were not optimal to carry out more complex experiments [8,9]. Type 1B astrocytes express constitutively the intermediate filaments, GFAP and vimentin, and NCAM 140, a cell surface molecule, in vivo and in vitro [18,19,24]. These cellular markers were used to characterize the cells by immunocytochemistry. Our technique used immunopanning to purify astrocytes derived from primary cultures derived from explanted human ONH. We used the anti-C5 monoclonal antibody, a neuroepithelial cell marker [13], as the first step of purification, on the premise that the most common contaminants (microglia, smooth muscle cells, lamina cribrosa cells, fibroblasts, meningeal cells, vascular) present in primary cultures would not attach to this epitope. The cells derived from C5 dish were highly pure GFAP+ astrocytes determined by immunocytochemistry and flow cytometry (f 95%). Cells derived from the second panning dish coated with the anti-Thy1.1 yielded cultures containing f60 – 75% GFAP+ astrocytes after removing contaminants under the microscope. From previous studies, type 1B astrocytes express constitutive GFAP in vivo and in vitro. Type 1B astrocytes also expressed intermediate filaments, vimentin and nestin as described previously. Type 1B astrocytes express vimentin in agreement with previous reports on astrocytes from rat optic nerve in vivo and in vitro [12]. Similarly, human and monkey type 1 B astrocytes express vimentin in vivo and retain expression in vitro [20]. Vimentin is an intermediate filament expressed by many nonneural cell types and astrocyte precursor cells (APC). Pax2 expression is specific to cells of the astrocyte lineage in the development of adult human retina and optic nerve head [3]. Pax2 is a transcription factor of the Pax family of DNA-binding proteins and may play a role in APC differentiation into astrocytes [12]. Heterozygous mutations of the Pax2 gene result in optic nerve coloboma in human and mice [3]. In our study, all GFAP+ astrocytes expressed Pax2 in the nucleus including astrocytes that expressed nestin derived from the Thy 1.1 dish. To determine that GFAP+ astrocytes derived from the anti-C5-coated dish were type 1B astrocytes, we performed immunocytochemistry for NCAM, a cell surface adhesion molecule also expressed on type 1B astrocytes in vivo and in vitro [9,24]. Our laboratory characterized expression of NCAM 140 in human type 1B astrocytes in detail [9,18,19]. NCAM is a cell surface adhesion molecule that may provide type 1B astrocyte with specialized properties in the ONH. We also reported that reactive astrocytes in the human optic nerve head changed expression of the spliced NCAM 140 isoform to the larger NCAM 180 in glaucomatous ONH [18]. Astrocytes derived from the C5 dish also expressed Thy1.1 in a similar pattern as NCAM. Thy1.1 and NCAM share the HNK1 epitope, a sulfated carbohydrate residue present in many
74
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
cell surface molecules thus explaining the cell surface staining [14]. Many cells (60 –75%) derived from the anti-Thy1.1coated dish expressed GFAP and vimentin. Twenty-five to thirty percent of these cells expressed nestin. These cells are indeed astrocytes. Nestin is an intermediate filament expressed in the optic nerve during development and by reactive astrocytes after injury [7]. It is possible that GFAP+/NCAM+ cells derived from remnant GFAP+ cells that did not adhere to the anti-C5 dish during the first panning, which subsequently attached to the anti-Thy1 dish by expression of the HNK1 epitope on cell surface molecules. Alternative, astrocytes GFAP+/NCAM+ derived from the Thy1.1 dish may represent a subpopulation of ONH astrocytes. Future experiments will test this possibility. In these cultures, GFAP+ and vimentin+ astrocytes that express nestin may be derived from immature astrocytes retained in adult optic nerve heads. Less than 5% of the cells derived from the Thy1.1 dish express SMA, but not GFAP or Pax2, suggesting smooth muscle cells or pericytes. A2B5-positive cells rarely grew from cells derived from either the anti-C5 or the anti Thy1.1-coated dish. A2B5 identifies O-A2 progenitor cells, type 2 astrocytes and oligodendrocytes present in the myelinated optic nerve of rodents and humans [6,24]. In our experiments, the myelinated nerve was removed from the explants under the phase microscope. In addition, our ONH culture conditions do not favor growth of oligodendrocytes [5]. In these experiments, microglia did not grow after immunopanning although they have been identified in the human optic nerve head by immunostaining with HLA DR [15]. Human microglial cells do not survive well after passage, nor can they withstand the time in culture required for astrocytes [5].
Under dissecting microscope, remove any pigmented and nonneural tissues. Identify optic nerve head (ONH) and cut it out. Bisect the ONH and remove the central vessel, cut tissues into four explants. 8.2. Explants culture
Prepare T25 flask with 0.5 ml culture medium. Rinse explants in culture medium and place into culture flask, let flask stand in the hood for 10 min. Add 0.5 ml culture medium and put flask into the incubator. Change medium twice a week, after the first cell appears, use 2 ml culture medium. 1 – 3 weeks later, remove explants and change medium to astrocytes-defined serum-free medium. One week later, do immunopanning.
8.3. Immunopanning 8.3.1. Prepare immunopanning dishes
Incubate two 10-mm Petri dishes overnight at 4 jC in 10 ml Tris (50 mM, pH 9.5) containing affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml). Wash dishes with 10 ml PBS, three times, 5 min each. Add 10 ml primary antibody in PBS/BSA (0.2%) and keep dishes in the hood at room temperature (RT) for 1.5 h. Coat one dish with C5 monoclonal antibody (1:20) and the other coat with Thy1.1 (1:1000). Wash dishes with 10 ml PBS, three times, 5 min each. Add 10 ml PBS to each dish and remove PBS right before use.
8.3.2. Prepare cell suspension 7. Conclusions
In conclusion, the immunopanning technique described here allows culture of highly purified type 1B astrocytes by immunopanning with anti C5 monoclonal antibody from normal human ONH. This method is suitable for small tissue samples like the human ONH, allows growing astrocytes from individual ONH without pooling from multiple donors, and does not require enzymatic dissociation of ONH tissue.
8.3.3. Immunopanning
8. Quick procedures 8.1. Dissection
Rinse eyes with sterile cold HBSS. Free optic nerve and transfer optic nerve to 35-mm dishes with 2 ml HBSS/10% dispase.
Remove medium from T25 flask, rinse with PBS. Add 4 ml cell suspension solution, keep in the incubator for 40 min. Pipet up and down to detach cells. Transfer cell suspension from same eye to 50-ml tube. Centrifuge at 4 jC, 1100 rpm for 5 min. Resuspend cells in 10 ml culture medium and filter through 70-Am cell strainer.
Add cell suspension to the C5-coated dish for 30 min at RT. Agitate C5-coated dish gently at 15 min. Transfer nonadherent cells from C5 dish to Thy1.1coated dish for 30 min. Wash C5 dish with DMEM/HEPES for three to five times, check process under microscope, add 10 ml DMEM/F-12/10% FBS to C5 dish and keep cell in the incubator.
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Agitate Thy1.1-coated dish gently at 15 min. Wash Thy1.1 dish with DMEM/HEPES for three to five times, check process under microscope, add 10 ml DMEM/F-12/10% FBS to Thy1.1 dish and keep cells in the incubator. Transfer nonadherent cells from Thy1.1 dish to an uncoated 100-mm Petri dish and keep cells in the incubator.
75
References
8.4. Flow cytometry
Trypsinize cells from T75 flasks for 12 min, spin at 4 jC, 1100 rpm for 5 min. Wash pellets by PBS, three times, 5 min each. Fix cells by 3 ml 4% paraformaldehyde for 30 min at RT. Wash pellets by PBS, three times, 5 min each. Permeabilize cells in PBS/5% FBS/0.2% Triton X-100/ 0.5% glycine for 30 min. Divide cells into three 15-ml centrifuge tubes, #1 add 150 Al 1j Ab rb anti-hu GFAP (1:100) for 45 min at RT. #2 and #3, no 1j Ab, just add antibody buffer. Wash with PBS, three times, 5 min each. Add 150 Al 2j Ab goat FITC-conjugated anti-rabbit IgG (1:200) to #1 and #3 for 45 min at RT. Only add 2j Ab buffer to #2. Wash with PBS, three times, 5 min each, store at dark until analysis. Measure the stained cells by FACScan flow cytometer.
8.5. Immunohistochemistry
Wash cells with cold PBS once for 5 min. Fix cells with 4% paraformaldehyde for 10 min. Wash with cold PBS, three times, 5 min each. For cytoskeleton staining, permeablize cells in 0.1% Triton X-100 for 10 min. Blocking by 100 Al 10% donkey serum in PBS/0.5% BSA for 30 min. Incubate with 1j Ab, wash and incubate with 2j Ab. (Table 1). Wash with PBS, three times, and mount with Vectashield. Examine slides under microscope and take pictures.
9. Essential literature references Refs. [2,9,12] Acknowledgements We would like to thank Dr. Ben Barres for providing C5 anti-neuroepithelial monoclonal antibody and his own protocols, and Dr. Junjie Yang for his technical help on flow cytometry.
[1] O.A. Agapova, C.S. Ricard, M. Salvador-Silva, M.R. Hernandez, Expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in human optic nerve head astrocytes, Glia 33 (2001) 205 – 216. [2] B.A. Barres, B.E. Silverstein, D.P. Corey, L.L. Chun, Immunological, morphological, and electrophysiological variation among retinal ganglion cells purified by panning, Neuron 1 (1988) 791 – 803. [3] Y. Chu, S. Hughes, T. Chan-Ling, Differentiation and migration of astrocyte precursor cells and astrocytes in human fetal retina: relevance to optic nerve coloboma, FASEB J. 15 (2001) 2013 – 2015. [4] I. Dabin, C.J. Barnstable, Rat retinal Muller cells express Thy-1 following neuronal cell death, Glia 14 (1995) 23 – 32. [5] C.J. De Groot, L. Montagne, I. Janssen, R. Ravid, D. Van, V.R. Veerhuis, Isolation and characterization of adult microglial cells and oligodendrocytes derived from postmortem human brain tissue, Brain Res. Protoc. 5 (2000) 85 – 94. [6] C. Ffrench-Constant, M.C. Raff, The oligodendrocyte-type-2 astrocyte cell lineage is specialized for myelination, Nature 323 (1986) 335 – 338. [7] J. Frisen, C.B. Johansson, C. Torok, M. Risling, U. Lendahl, Rapid, widespread, and longlasting induction of nestin contributes to the generation of glial scar tissue after CNS injury, J. Cell Biol. 131 (1995) 453 – 464. [8] M.R. Hernandez, F. Igoe, A.H. Neufeld, Cell culture of the human lamina cribrosa, Investig. Ophthalmol. Vis. Sci. 29 (1988) 78 – 89. [9] S. Kobayashi, I. Vidal, J.D. Pena, M.R. Hernandez, Expression of neural cell adhesion molecule (NCAM) characterizes a subpopulation of type 1 astrocytes in human optic nerve head, Glia 20 (1997) 262 – 273. [10] B. Liu, A.H. Neufeld, Expression of nitric oxide synthase-2 (NOS-2) in reactive astrocytes of the human glaucomatous optic nerve head, Glia 30 (2000) 178 – 186. [11] B. Liu, A.H. Neufeld, Nitric oxide synthase-2 in human optic nerve head astrocytes induced by elevated pressure in vitro, Arch. Ophthalmol. 119 (2001) 240 – 245. [12] H. Mi, B.A. Barres, Purification and characterization of astrocyte precursor cells in the developing rat optic nerve, J. Neurosci. 19 (1999) 1049 – 1061. [13] R.H. Miller, B.P. Williams, J. Cohen, M.C. Raff, A4: an antigenic marker of neural tube-derived cells, J. Neurocytol. 13 (1984) 329 – 338. [14] J.R. Naegele, C.J. Barnstable, A carbohydrate epitope defined by monoclonal antibody VC1.1 is found on N-CAM and other cell adhesion molecules, Brain Res. 559 (1991) 118 – 129. [15] A.H. Neufeld, Microglia in the optic nerve head and the region of parapapillary chorioretinal atrophy in glaucoma, Arch. Ophthalmol. 117 (1999) 1050 – 1056. [16] J.D. Pena, O.A. Agapova, B.T. Gabelt, L.A. Levin, M.J. Lucarelli, P.L. Kaufman, M.R. Hernandez, Increased elastin expression in astrocytes of the lamina cribrosa in response to elevated intraocular pressure, Investig. Ophthalmol. Vis. Sci. 42 (2001) 2303 – 2314. [17] G. Prasanna, R. Krishnamoorthy, A.F. Clark, R.J. Wordinger, T. Yorio, Human optic nerve head astrocytes as a target for endothelin-1, Investig. Ophthalmol. Vis. Sci. 43 (2002) 2704 – 2713. [18] C.S. Ricard, J.D. Pena, M.R. Hernandez, Differential expression of neural cell adhesion molecule isoforms in normal and glaucomatous human optic nerve heads, Mol. Brain Res. 74 (1999) 69 – 82. [19] C.S. Ricard, S. Kobayashi, J.D. Pena, M. Salvador-Silva, O.A. Agapova, M.R. Hernandez, Selective expression of neural cell adhesion molecule (NCAM)-180 in optic nerve head astrocytes exposed to elevated hydrostatic pressure in vitro, Mol. Brain Res. 81 (2000) 62 – 79. [20] M. Salvador-Silva, C.S. Ricard, O.A. Agapova, P. Yang, M.R. Hernandez, Expression of small heat shock proteins and intermediate filaments in the human optic nerve head astrocytes exposed to
76
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
elevated hydrostatic pressure in vitro, J. Neurosci. Res. 66 (2001) 59 – 73. [21] G. Tezel, M.B. Wax, Increased production of tumor necrosis factoralpha by glial cells exposed to simulated ischemia or elevated hydrostatic pressure induces apoptosis in cocultured retinal ganglion cells, J. Neurosci. 20 (2000) 8693 – 8700. [22] M.B. Wax, G. Tezel, Neurobiology of glaucomatous optic neuropathy: diverse cellular events in neurodegeneration and neuroprotection, Mol. Neurobiol. 26 (2002) 45 – 55.
[23] R.J. Wordinger, R. Agarwal, M. Talati, J. Fuller, W. Lambert, A.F. Clark, Expression of bone morphogenetic proteins (BMP), BMP receptors, and BMP associated proteins in human trabecular meshwork and optic nerve head cells and tissues, Mol.Vis. 8 (2002) 241 – 250. [24] H. Ye, M.R. Hernandez, Heterogeneity of astrocytes in human optic nerve head, J. Comp. Neurol. 362 (1995) 441 – 452.
Protocols
Purification of astrocytes from adult human optic nerve heads by immunopanning Ping Yang a, M. Rosario Hernandez a,b,* a
Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, St. Louis, MO, USA b Department of Anatomy and Neurobiology, Washington University School of Medicine, St. Louis, MO, USA Accepted 16 July 2003
Abstract Most in vitro studies in the CNS require pure cultures of astrocytes. Astrocytes from the human optic nerve head (ONH, type 1B) represent a specialized population of astrocytes. Primary cells grown from human optic nerve head explants were cultured for 3 – 4 weeks. To select astrocytes by immunopanning, cell suspensions were placed on a P100 panning dish coated with C5 anti-neuroepithelial antibody and allowed to attach for 30 min. Nonadherent cells were plated on a second dish coated with anti-Thy1.1 antibody to deplete microglia and meningeal cells. Finally, remnant nonadherent cells were plated on a noncoated dish. Purified cells were immunostained with astrocyte markers: GFAP, vimentin, Pax2, A2B5, nestin and NCAM. Other cell types were characterized by HLA-DR for microglia and smooth muscle actin for vascular smooth muscle. The proportion of GFAP+ astrocytes in the cultures was determined by flow cytometry. About 95% of the cells that adhered to the C5 dish were GFAP+ astrocytes. GFAP+ astrocytes expressed vimentin, Pax2, nestin and NCAM, but not A2B5. From the Thy1.1 dish, 60 – 75% cells were GFAP+ astrocytes and the remainder cells were GFAP cells. Using cloning rings, we eliminated fibroblast-like cells, smooth muscle and meningeal cells from astrocyte cultures. Smooth muscle cells and fibroblasts grew on the noncoated dish. In conclusion, immunopanning is an efficient method to get high yields of viable type 1B astrocytes from adult human ONH. The current described culture system may provide a valuable tool in studying human optic nerve head biology and disease. D 2003 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Astrocyte, optic nerve head Keywords: Human optic nerve head; Astrocyte; Explants culture
1. Type of research
Astrocytes are the major glial cell type in the nonmyelinated optic nerve head (ONH) in most mammals and provide cellular support functions to the axons of the retinal ganglion cells while interfacing between connective tissue surfaces and surrounding blood vessels. In the ONH, transition from quiescent to reactive astrocyte occurs in response to various forms of stress and in disease such as proinflammatory cytokines, injury, and in degenerative diseases such as glaucomatous optic neuropathy.
To study cellular responses of ONH astrocytes in physiologic or pathophysiological conditions, we describe here a culture system to obtain purified OHN astrocytes (type 1B) from human postmortem eyes by immunopanning. Characterization of astrocyte cultures by immunocytochemistry and flow cytometry.
2. Time required
* Corresponding author. Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, 660 S. Euclid Avenue, Campus Box 8096, St. Louis, MO 63110, USA. Tel.: +1-314-7471448; fax: +1-314-747-1405. E-mail address: [email protected] (M.R. Hernandez). 1385-299X/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S1385-299X(03)00073-4
Preparation of reagents, 1.5 h Dissecting ONH and explant preparation: 2 h Explant culture: 1 –3 weeks Preparation of immunopanning dishes: overnight Dissociation of cells from explant culture: 40 min Panning procedure: 1.5 h Culture of cells in panning dishes: 1 –2 weeks Trypsinization and expansion of cultures: 1 –2 weeks
68
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Immunocytochemistry: 4 h Flow cytometry: 4 h Freezing stock of cells for further use: 2 h.
3. Materials
3.1. Human eye tissue
Human eyes without history of eye disease, neurodegenerative disease or diabetes (age 15 – 46) were obtained within 24 h of death from Mid-America Transplant Services (St. Louis, MO) and National Disease Research Interchange (NDRI, Philadelphia, PA). No donor consent was required for eyes used in research.
3.2. Special equipment
Culture dishes (100 and 35 mm; Falcon, Lincoln Park, NJ) six-well plates (Falcon) Primaria Tissue culture flasks (25 cm2, Falcon) Primaria Tissue culture flasks (75 cm2, Falcon) Pipettes (10, 5, 2 ml, Fisher Scientific, Pittsburgh, PA) Cover glasses (22 mm, Fisher Scientific) Stericup filter (Millipore, Fisher Scientific) Centrifuge tubes (50 ml, Fisher Scientific) Cell strainer (70 Am, Becton Dickinson, Franklin Lakes, NJ) Forceps (large, small) (Storz, St. Louis, MO) Surgical scissors (Storz) Surgical blades, 15 mm (Feather Safety Razor, Japan) Dissection microscope (Carl Zeiss, Thornwood, NY) Phase-contrast reversal microscope (Olympus CK2, Japan) TMAX black and white film (Kodak, Rochester, NY) Nikon Optiphot-2 microscope (Tokyo, Japan) Digital camera (Spot Diagnostic Instruments, Sterling Heights, MI) FACScan flow cytometer/CELLQuest Software system (Becton-Dickinson, San Jose, CA).
3.3. Chemicals and reagents
Hank’s balanced salt solution (HBSS, Washington University Medical School Tissue Culture Support Center, St. Louis, MO) Dispase (Sigma, St. Louis, MO) Dulbecco’s modified Eagle’s medium (DMEM)/F-12 (Washington University Medical School Tissue Culture Support Center) Fetal Bovine Serum (FBS, Biowhittaker, Walkerswille, MD) PSFM (10,000 U/ml penicillin, 10,000 Ag/ml streptomycin and 25 Ag/ml amphotericin B) (Gibco/BRL, Gaithersburg, MD)
1.3% bovine albumin fraction V (Sigma) ITS+ Premixk culture supplement (Collaborative Biomedical Products, Bedford, MA) Affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml, Jackson ImmunoResearch Lab, West Grove, PA) C5 neuroepithelial monoclonal antibody (1:20, gift from Dr. Ben Barres, Stanford University) Thy1.1 monoclonal antibody (1:1000, Chemicon, Temecula, CA) Cell suspension solution (Sigma) Rabbit anti-human GFAP (1:100, Sigma) Goat FITC-conjugated anti-rb IgG (1:200, Sigma) Donkey serum (Sigma) Vectashield (Vector, Burlingame, CA) PBS (Invitrogen) BSA (Fisher) Tris (50 mM, pH 9.5, Fisher) Alexa Fluork goat anti-rabbit or anti-mouse IgG (H + L) 488 (1:600, Molecular Probes, Eugene, OR) Alexa Fluork goat anti-rabbit or anti-mouse IgG (H + L) 568 (1:800, Molecular Probes) Texas Red-conjugated anti-mouse IgM (1:100, Chemicon) FITC-conjugated anti-mouse IgM (1:50, Chemicon).
4. Detailed procedures 4.1. Dissection Ten normal human eyes without history of eye disease, diabetes or chronic neurodegenerative disease (age 15– 46) were obtained within 24 h of death from eye banks throughout the United States through National Disease Research Interchange (NDRI) and the Mid-America Transplant Services. Eyes were transported to the laboratory in special sterile containers in a humidified atmosphere on ice, placed in sterile 100-mm Petri dishes and carefully rinsed with sterile cold 10-ml Hank’s balanced salt solution. The optic nerve head was freed from sclera and other neighboring tissues using surgical scissors and forceps and then transferred to a 35-mm Petri dish with 2 ml HBSS/10% Dispase. Under a dissecting microscope, pigmented and remnant nonneural tissues were removed and the post laminar myelinated nerve was identified and removed using a sharp blade. The resulting cylinder of tissue was bisected and the central vessels removed. The tissues were then cut into four explants. 4.2. Explants culture Tissue culture flasks (25 cm2) were conditioned with 0.5 ml growth medium: Dulbecco’s modified Eagle’s medium (DMEM)/F-12 supplemented with 10% FBS and PSFM (10,000 U/ml penicillin, 10,000 Ag/ml streptomycin and 25 Ag/ml amphotericin B), and prewarmed in a 37 jC, 5% CO2 incubator. Each explant was rinsed once in growth
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
medium, placed into the culture flask, and kept standing in the hood for 10 min until the explant adhered to the surface of the flask. Before placing in the incubator, 0.5 ml DMEM/ F-12/10% FBS was added to each flask. The growth medium was changed twice a week. When the first cells began to grow out of the explant, the medium was increased to 2 ml. One to three weeks later, the explants were removed and the cultures were fed with astrocyte-defined serum-free medium containing DMEM, 1.3% bovine albumin fraction V, 1 Al/ml ITS+ Premixk culture supplement and PSFM. Cells were allowed to grow to cover an area around the explants. Approximately, 10,000 cells are obtained per explant. Cells from three or four explants (30 – 50,000 cells) from the same eye were combined for panning. 4.3. Immunopanning Immunopanning was performed about one week after removing the explants. The immunopanning dishes were prepared the day before panning as described in detail by Mi and Barres [12]. Two 100-mm Petri dishes were incubated overnight at 4 jC in 10 ml Tris (50 mM, pH 9.5) containing affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml). The next day, each dish was washed three times with PBS. Following wash, 10 ml primary antibody in PBS/BSA (0.2%) was added and incubated at RT for 1.5 h. One dish was incubated with C5 neuroepithelial monoclonal antibody (1:20, gift from Dr. Ben Barres, Stanford University) and the second dish with Thy1.1 monoclonal antibody (1:1000). Afterwards, the dishes were rinsed with PBS, three times. Finally, 10 ml PBS was added to coat the dishes. Remove the PBS just before use. A cell suspension was prepared after removing media from T25 flasks. Briefly, 4 ml cell suspension solution was added to each flask and placed in the incubator for 40 min, then detach the cells by pipetting them up and down. The cell suspension was transferred to a 50-ml centrifuge tube and the cells spun down at 1100 rpm. The cells were then resuspended in 10 ml DMEM/F-12/10% FBS and filtered through a 70-Am pore cell strainer. A 200-Al aliquot was used to count cells and the rest of the cell suspension was incubated on the C5-coated dish for 30 min at room temperature in the hood with gentle agitation at 15 min to ensure access of all cells to the panning surface area. Next, the nonadherent cells were transferred to the Thy1.1-coated dish for 30 min. The C5 dish was gently washed with DMEM/HEPES for three to five times to remove the nonadherent cells. This step was monitored under the inverted microscope to make sure cells do not dry out and only the nonadherent cells were removed. The times of wash are important. Too many washes will wash off astrocytes, too few washes will not wash off the nonadherent cells completely. Afterwards, 10 ml DMEM/F-12/10% FBS medium was added to the C5 dish and placed in the incubator. The procedure with the Thy1.1 dish was the same. After panning, cells were
69
allowed to reach confluence in 1 – 2 weeks. Clusters of cells that were obvious contaminants were removed at this point using trypsinization (0.05% trypsin for 10 min) and cloning rings under the microscope. The purified cultures were then expanded by passing into T75 flasks. Samples of each culture were plated on coverslips for characterization by immunocytochemistry. The morphology of live cells was observed using an Olympus CK2 phase-contrast microscope. Images were recorded using photographic camera (Nikon NFX-35) using TMAX 400, black and white film. 4.4. Flow cytometry For flow cytometry, cells were grown in 75-cm2 flasks until confluence. Cells were then trypsinized and spun down at 4 jC. Cell pellets were washed three times by PBS. After removing PBS, the cells were fixed by adding 3 ml 4% paraformaldehyde for 30 min at room temperature. After two washes in PBS, cells were permeabilized in PBS containing 5% FBS, 0.2% Triton X-100 and 0.5% glycine for 30 min. The cells were divided into three 15-ml centrifuge tubes (f 2 106 cells per tube), one for sample (primary and secondary antibody present), one control for autofluorescence (without primary and secondary antibodies) and one for negative control (only secondary antibody). The sample was incubated with 150 Al primary antibody (rabbit anti-human GFAP 1:100) in PBS/1% BSA/0.5% Triton X-100 at room temperature for 45 min. After washing with PBS for three times, 150 Al secondary antibody, goat FITC-conjugated anti-rabbit IgG (1:200) in PBS/1% BSA was added to the sample and negative control. Cells were incubated in secondary antibody for
Table 1 Primary and secondary antibodies for immunocytochemistry Primary antibodies
1j Ab Source
1j Ab Dilution
Secondary antibodies
Mouse anti-human GFAP Mouse anti-human Vimentin Rabbit anti-human Nestin Mouse anti-human SMA Rabbit anti-human Pax2 Mouse anti-human NCAM Mouse anti-human A2B5 Mouse anti-human HLA-DR Mouse anti-human Thy1.1
Sigma
1:50
Sigma
1:50
Gift, Dr. Lundkvist Biogenex
1:1000 1:30
BabCO
1:50
Accurate
1:50
Chemicon
1:200
Accurate
1:50
Chemicon
1:200
Alexa goat anti-mouse IgG (H + L) 568 (1:800) Texas Red-conjugated anti-mouse IgM (1:100) Alexa goat anti-rabbit IgG (H + L) 488 (1:600) Alexa goat anti-mouse IgG (H + L) 568 (1:800) Alexa goat anti-rabbit IgG (H + L) 488 (1:600) Texas Red-conjugated anti-mouse IgM (1:100) FITC-conjugated anti-mouse IgM (1:50) Alexa goat anti-mouse IgG (H + L) 568 (1:800) Alexa goat anti-mouse IgG (H + L) 568 (1:800)
70
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
4.5. Immunocytochemistry For characterization, cells were cultured in 6 35-mm well plates containing 22-mm glass coverslips in growth medium until confluence. Cells were characterized by staining with astrocyte markers: glial fibrillary acidic protein (GFAP), vimentin, Pax2, nestin, neural cell adhesion molecule (NCAM) and A2B5. Other markers used included HLA-DR for microglia and smooth muscle actin (SMA) for vascular smooth muscle. Primary and secondary antibodies are summarized in Table 1. All primary antibodies were diluted in a blocking agent (10% donkey serum in PBS/ 0.5% BSA). All secondary antibodies were prepared in PBS/0.5% BSA.
Fig. 1. Phase-contrast micrographs of live primary cultured cells (age 15) from optic nerve head. (A) Cells grown on the C5 panning dishes have starshaped morphology. (B) Most of the cells grown on the Thy1.1 panning dishes have the same morphology as A, arrows point to large cells represent most likely lamina cribrosa cells. (C) Nonadherent cells were thin, elongated fibroblast-like cells or smooth muscle cells. Objective: A and B, 10 ; C, 20 .
45 min at room temperature. After washing, stained cells were maintained in the dark until analysis. The stained cells were measured using a FACScan flow cytometer/CELLQuest Software system. A single, aircooled, argon-ion laser provided excitation. Data were collected using logarithmic amplification on 5000 cells, excluding cell debris by a combination of forward and side scatters. Frequency histograms representative of cells obtained after panning with C5 and Thy1.1 are shown (Fig. 2). At least 10 cultures generated by panning from different donors have been tested.
Fig. 2. Flow cytometric analysis of cells adherent to C5 and Thy1.1 panning dishes. Percentage of GFAP-positive cells (after subtracting for nonspecific immunofluorescence, dotted line) adherent to C5 (A) and Thy1.1 (B) panning dishes.
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Cells on coverslips were washed with cold PBS once and fixed with cold 4% paraformaldehyde for 10 min, then washed three times by adding 1.5 ml of ice-cold PBS for 5 min. For cytoskeleton staining, cells were permeabilized in 0.1% Triton X-100 for 10 min. After blocking by incubation with 100 Al 10% donkey serum in PBS/ 0.5% BSA for 30 min, 100 Al of primary antibody was added to the coverslips and incubated at room temperature for 1 h. Cells were then washed three times in PBS and incubated with 100 Al appropriate secondary antibody for 45 min. After washing three times in PBS, the coverslips were mounted with Vectashield. For control studies, the primary antibody was replaced with nonimmune serum or omitted. Fluorescent images of immunolabeled cells were obtained using a Nikon Optiphot-2 microscope equipped with appropriate fluorescence filters for single or double fluorescence. Images were recorded using a digital camera and stored as computer files.
71
5. Results 5.1. Primary cell culture from human optic nerve head Initial outgrowth from the explanted lamina cribrosa appeared within 1 – 3 weeks in culture. Cells formed a monolayer around the explant and consisted of a mixture of flat polygonal cells with and without processes, and large flat cells with oval clear nuclei. Fibroblast-like cells formed independent clusters of elongated cells. 5.2. Immunopanning To purify astrocytes from explant culture, we used and modified the sequential panning procedure reported originally by Mi and Barres [12] and Barres et al. [2]. Briefly, the cell suspension of primary cells was panned first on a Petri dish coated with C5 anti-neuroepithelial monoclonal antibody to select for cells of astrocyte lineage. The nonadher-
Fig. 3. Immunofluoresence staining of cells (age 46) adherent to C5 panning dish. Positive intracellular staining of intermediate filaments of GFAP (A), vimentin (B) and nestin (C). Positive cell surface staining of NCAM (D) and Thy1.1 (E). Positive nuclear staining of Pax2 (G). Negative staining of A2B5 (F), SMA (H) and HLA-DR (I).
72
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
ent cells from the C5 dish were then passed onto a dish coated with the anti-Thy1.1 monoclonal antibody, a cell surface molecule member of the immunoglobulin superfamily [4] to deplete microglia, fibroblasts and meningeal cells. The nonadherent cells from the Thy1.1 dish were discarded. Almost all cells attached to the C5-coated dish were starshaped flat cells, with short processes. Very few other contaminant cell types were present (Fig 1A). Once the cells attached to the anti-C5-coated dish reached confluence, cells were detached by trypsinization and then plated on coverslips for characterization by immunocytochemistry and to be expanded for flow cytometry testing and further use in the laboratory. Interestingly, many cells attached to the anti-Thy1.1coated dish were also star-shaped cells, similar to the C5 dish (Fig 1B). Clusters of flat, small polygonal cells and isolated large cells with large nuclei represent most likely lamina cribrosa cells and meningeal cells, respectively. These cells were morphologically distinct from astrocytes
and they were removed from the culture dish using cloning rings and trypsinization (0.05% trypsin for 10 min). Once cells attached to the anti-Thy1.1 dish reached confluence, they were plated on coverslips for further characterization. The nonadherent cells from the anti-Thy1.1 dish were thin, elongated, fibroblast-like cells or smooth muscle cells (Fig 1C). These cells were discarded. 5.3. Flow cytometry In order to evaluate the purity of the astrocyte cultures derived from the anti-C5 and the anti-Thy1.1 dishes we performed flow cytometry. Astrocytes were cultured to confluence in astrocyte-defined serum-free medium for at least 10 days before flow cytometry. After labeling with GFAP and fluorescein, at least 95% cells derived from the anti-C5 dish were GFAP-positive astrocytes. Sixty to seventy-five percent of the cells derived from the anti-Thy1.1coated dish were also astrocytes (Fig 2).
Fig. 4. Immunofluoresence staining of cells (age 46) adherent to Thy1.1 panning dish. Positive intracellular staining of intermediate filaments of GFAP (A), vimentin (B) and nestin (C). Positive cell surface staining of NCAM (D) and Thy1.1 (E). Positive nuclear staining of Pax2 (G). Few cells have positive staining of A2B5 (F) and SMA (H). Negative staining of HLA-DR (I).
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
5.4. Immunocytochemistry The cytoskeleton of cells derived from the anti-C5coated dish stained for the intermediate filaments, GFAP, vimentin and nestin (Fig. 3). More than 95% cells derived from the anti-C5-coated dish were GFAP- and vimentinpositive. Few GFAP+ astrocytes (less than 10%) derived from the anti-C5 dish expressed nestin. Pax2, a transcription factor expressed by astrocyte lineage cells, labeled the nuclei of almost all cells derived from the anti-C5 dish. HLA-DR, a marker of microglia, was not expressed in cells derived from the anti-C5 dish. There were no cells that expressed smooth muscle actin, a marker of vascular smooth muscle. The cytoskeleton of 60 –75% cells derived from the antiThy1.1 dish stained with GFAP and vimentin (Fig. 4). Approximately 25 –35% of the cells expressed nestin and GFAP. All GFAP+ cells expressed Pax 2. Small clusters of cells expressed smooth muscle actin and they are presumably smooth muscle cells or pericytes. No microglial cell derived from the anti-Thy1.1 dish. NCAM is a cell surface adhesion molecule expressed by type 1B astrocytes of the human optic nerve head in vivo and in vitro [18,19]. Almost all GFAP+ cells derived from the C5 dish expressed NCAM as fine granules on the cell surface (Fig. 3). In addition, these cells reacted with the antiThy1.1 antibody in a similar pattern with anti-NCAM antibody. No cells derived from the anti-C5 dish expressed A2B5, a cell surface marker that labels type 2 astrocytes and precursors of the oligodendrocyte lineage. Between 50% and 75% of the GFAP+ cells derived from the Thy1.1 dish expressed NCAM and all cells expressed Thy1.1 (Fig 4). Few isolated GFAP+ cells derived from the Thy1.1 dish stained with the cell surface marker A2B5 (Fig. 4).
6. Discussion Here we describe a procedure to culture highly purified human optic nerve head astrocytes (type 1B) using a immunopanning method for rat optic nerve astrocytes modified from Mi and Barres [12] and Barres et al. [2]. Type 1B astrocytes are the major glial cell subpopulation in the nonmyelinated optic nerve head [24]. The ONH is thought to be the site of retinal ganglion cell (RGC) damage in primary open angle glaucoma (POAG), a common blinding neurodegenerative disease [22]. In glaucoma, there is extensive remodeling of the extracellular matrix (ECM) of the ONH with increased expression of ECM and proteolytic enzymes, cell adhesion molecules and neurotoxic mediators by reactive ONH astrocytes [1,11,16,21]. Therefore, ONH astrocytes represent a possible target for neuroprotection in glaucomatous optic neuropathy [10,17,23].Our laboratory developed techniques to culture type 1B astrocytes, but the yield, purity and viability of the cultures after multiple
73
passages were not optimal to carry out more complex experiments [8,9]. Type 1B astrocytes express constitutively the intermediate filaments, GFAP and vimentin, and NCAM 140, a cell surface molecule, in vivo and in vitro [18,19,24]. These cellular markers were used to characterize the cells by immunocytochemistry. Our technique used immunopanning to purify astrocytes derived from primary cultures derived from explanted human ONH. We used the anti-C5 monoclonal antibody, a neuroepithelial cell marker [13], as the first step of purification, on the premise that the most common contaminants (microglia, smooth muscle cells, lamina cribrosa cells, fibroblasts, meningeal cells, vascular) present in primary cultures would not attach to this epitope. The cells derived from C5 dish were highly pure GFAP+ astrocytes determined by immunocytochemistry and flow cytometry (f 95%). Cells derived from the second panning dish coated with the anti-Thy1.1 yielded cultures containing f60 – 75% GFAP+ astrocytes after removing contaminants under the microscope. From previous studies, type 1B astrocytes express constitutive GFAP in vivo and in vitro. Type 1B astrocytes also expressed intermediate filaments, vimentin and nestin as described previously. Type 1B astrocytes express vimentin in agreement with previous reports on astrocytes from rat optic nerve in vivo and in vitro [12]. Similarly, human and monkey type 1 B astrocytes express vimentin in vivo and retain expression in vitro [20]. Vimentin is an intermediate filament expressed by many nonneural cell types and astrocyte precursor cells (APC). Pax2 expression is specific to cells of the astrocyte lineage in the development of adult human retina and optic nerve head [3]. Pax2 is a transcription factor of the Pax family of DNA-binding proteins and may play a role in APC differentiation into astrocytes [12]. Heterozygous mutations of the Pax2 gene result in optic nerve coloboma in human and mice [3]. In our study, all GFAP+ astrocytes expressed Pax2 in the nucleus including astrocytes that expressed nestin derived from the Thy 1.1 dish. To determine that GFAP+ astrocytes derived from the anti-C5-coated dish were type 1B astrocytes, we performed immunocytochemistry for NCAM, a cell surface adhesion molecule also expressed on type 1B astrocytes in vivo and in vitro [9,24]. Our laboratory characterized expression of NCAM 140 in human type 1B astrocytes in detail [9,18,19]. NCAM is a cell surface adhesion molecule that may provide type 1B astrocyte with specialized properties in the ONH. We also reported that reactive astrocytes in the human optic nerve head changed expression of the spliced NCAM 140 isoform to the larger NCAM 180 in glaucomatous ONH [18]. Astrocytes derived from the C5 dish also expressed Thy1.1 in a similar pattern as NCAM. Thy1.1 and NCAM share the HNK1 epitope, a sulfated carbohydrate residue present in many
74
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
cell surface molecules thus explaining the cell surface staining [14]. Many cells (60 –75%) derived from the anti-Thy1.1coated dish expressed GFAP and vimentin. Twenty-five to thirty percent of these cells expressed nestin. These cells are indeed astrocytes. Nestin is an intermediate filament expressed in the optic nerve during development and by reactive astrocytes after injury [7]. It is possible that GFAP+/NCAM+ cells derived from remnant GFAP+ cells that did not adhere to the anti-C5 dish during the first panning, which subsequently attached to the anti-Thy1 dish by expression of the HNK1 epitope on cell surface molecules. Alternative, astrocytes GFAP+/NCAM+ derived from the Thy1.1 dish may represent a subpopulation of ONH astrocytes. Future experiments will test this possibility. In these cultures, GFAP+ and vimentin+ astrocytes that express nestin may be derived from immature astrocytes retained in adult optic nerve heads. Less than 5% of the cells derived from the Thy1.1 dish express SMA, but not GFAP or Pax2, suggesting smooth muscle cells or pericytes. A2B5-positive cells rarely grew from cells derived from either the anti-C5 or the anti Thy1.1-coated dish. A2B5 identifies O-A2 progenitor cells, type 2 astrocytes and oligodendrocytes present in the myelinated optic nerve of rodents and humans [6,24]. In our experiments, the myelinated nerve was removed from the explants under the phase microscope. In addition, our ONH culture conditions do not favor growth of oligodendrocytes [5]. In these experiments, microglia did not grow after immunopanning although they have been identified in the human optic nerve head by immunostaining with HLA DR [15]. Human microglial cells do not survive well after passage, nor can they withstand the time in culture required for astrocytes [5].
Under dissecting microscope, remove any pigmented and nonneural tissues. Identify optic nerve head (ONH) and cut it out. Bisect the ONH and remove the central vessel, cut tissues into four explants. 8.2. Explants culture
Prepare T25 flask with 0.5 ml culture medium. Rinse explants in culture medium and place into culture flask, let flask stand in the hood for 10 min. Add 0.5 ml culture medium and put flask into the incubator. Change medium twice a week, after the first cell appears, use 2 ml culture medium. 1 – 3 weeks later, remove explants and change medium to astrocytes-defined serum-free medium. One week later, do immunopanning.
8.3. Immunopanning 8.3.1. Prepare immunopanning dishes
Incubate two 10-mm Petri dishes overnight at 4 jC in 10 ml Tris (50 mM, pH 9.5) containing affinity-purified goat anti-mouse IgG (H + L) (10 Ag/ml). Wash dishes with 10 ml PBS, three times, 5 min each. Add 10 ml primary antibody in PBS/BSA (0.2%) and keep dishes in the hood at room temperature (RT) for 1.5 h. Coat one dish with C5 monoclonal antibody (1:20) and the other coat with Thy1.1 (1:1000). Wash dishes with 10 ml PBS, three times, 5 min each. Add 10 ml PBS to each dish and remove PBS right before use.
8.3.2. Prepare cell suspension 7. Conclusions
In conclusion, the immunopanning technique described here allows culture of highly purified type 1B astrocytes by immunopanning with anti C5 monoclonal antibody from normal human ONH. This method is suitable for small tissue samples like the human ONH, allows growing astrocytes from individual ONH without pooling from multiple donors, and does not require enzymatic dissociation of ONH tissue.
8.3.3. Immunopanning
8. Quick procedures 8.1. Dissection
Rinse eyes with sterile cold HBSS. Free optic nerve and transfer optic nerve to 35-mm dishes with 2 ml HBSS/10% dispase.
Remove medium from T25 flask, rinse with PBS. Add 4 ml cell suspension solution, keep in the incubator for 40 min. Pipet up and down to detach cells. Transfer cell suspension from same eye to 50-ml tube. Centrifuge at 4 jC, 1100 rpm for 5 min. Resuspend cells in 10 ml culture medium and filter through 70-Am cell strainer.
Add cell suspension to the C5-coated dish for 30 min at RT. Agitate C5-coated dish gently at 15 min. Transfer nonadherent cells from C5 dish to Thy1.1coated dish for 30 min. Wash C5 dish with DMEM/HEPES for three to five times, check process under microscope, add 10 ml DMEM/F-12/10% FBS to C5 dish and keep cell in the incubator.
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
Agitate Thy1.1-coated dish gently at 15 min. Wash Thy1.1 dish with DMEM/HEPES for three to five times, check process under microscope, add 10 ml DMEM/F-12/10% FBS to Thy1.1 dish and keep cells in the incubator. Transfer nonadherent cells from Thy1.1 dish to an uncoated 100-mm Petri dish and keep cells in the incubator.
75
References
8.4. Flow cytometry
Trypsinize cells from T75 flasks for 12 min, spin at 4 jC, 1100 rpm for 5 min. Wash pellets by PBS, three times, 5 min each. Fix cells by 3 ml 4% paraformaldehyde for 30 min at RT. Wash pellets by PBS, three times, 5 min each. Permeabilize cells in PBS/5% FBS/0.2% Triton X-100/ 0.5% glycine for 30 min. Divide cells into three 15-ml centrifuge tubes, #1 add 150 Al 1j Ab rb anti-hu GFAP (1:100) for 45 min at RT. #2 and #3, no 1j Ab, just add antibody buffer. Wash with PBS, three times, 5 min each. Add 150 Al 2j Ab goat FITC-conjugated anti-rabbit IgG (1:200) to #1 and #3 for 45 min at RT. Only add 2j Ab buffer to #2. Wash with PBS, three times, 5 min each, store at dark until analysis. Measure the stained cells by FACScan flow cytometer.
8.5. Immunohistochemistry
Wash cells with cold PBS once for 5 min. Fix cells with 4% paraformaldehyde for 10 min. Wash with cold PBS, three times, 5 min each. For cytoskeleton staining, permeablize cells in 0.1% Triton X-100 for 10 min. Blocking by 100 Al 10% donkey serum in PBS/0.5% BSA for 30 min. Incubate with 1j Ab, wash and incubate with 2j Ab. (Table 1). Wash with PBS, three times, and mount with Vectashield. Examine slides under microscope and take pictures.
9. Essential literature references Refs. [2,9,12] Acknowledgements We would like to thank Dr. Ben Barres for providing C5 anti-neuroepithelial monoclonal antibody and his own protocols, and Dr. Junjie Yang for his technical help on flow cytometry.
[1] O.A. Agapova, C.S. Ricard, M. Salvador-Silva, M.R. Hernandez, Expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in human optic nerve head astrocytes, Glia 33 (2001) 205 – 216. [2] B.A. Barres, B.E. Silverstein, D.P. Corey, L.L. Chun, Immunological, morphological, and electrophysiological variation among retinal ganglion cells purified by panning, Neuron 1 (1988) 791 – 803. [3] Y. Chu, S. Hughes, T. Chan-Ling, Differentiation and migration of astrocyte precursor cells and astrocytes in human fetal retina: relevance to optic nerve coloboma, FASEB J. 15 (2001) 2013 – 2015. [4] I. Dabin, C.J. Barnstable, Rat retinal Muller cells express Thy-1 following neuronal cell death, Glia 14 (1995) 23 – 32. [5] C.J. De Groot, L. Montagne, I. Janssen, R. Ravid, D. Van, V.R. Veerhuis, Isolation and characterization of adult microglial cells and oligodendrocytes derived from postmortem human brain tissue, Brain Res. Protoc. 5 (2000) 85 – 94. [6] C. Ffrench-Constant, M.C. Raff, The oligodendrocyte-type-2 astrocyte cell lineage is specialized for myelination, Nature 323 (1986) 335 – 338. [7] J. Frisen, C.B. Johansson, C. Torok, M. Risling, U. Lendahl, Rapid, widespread, and longlasting induction of nestin contributes to the generation of glial scar tissue after CNS injury, J. Cell Biol. 131 (1995) 453 – 464. [8] M.R. Hernandez, F. Igoe, A.H. Neufeld, Cell culture of the human lamina cribrosa, Investig. Ophthalmol. Vis. Sci. 29 (1988) 78 – 89. [9] S. Kobayashi, I. Vidal, J.D. Pena, M.R. Hernandez, Expression of neural cell adhesion molecule (NCAM) characterizes a subpopulation of type 1 astrocytes in human optic nerve head, Glia 20 (1997) 262 – 273. [10] B. Liu, A.H. Neufeld, Expression of nitric oxide synthase-2 (NOS-2) in reactive astrocytes of the human glaucomatous optic nerve head, Glia 30 (2000) 178 – 186. [11] B. Liu, A.H. Neufeld, Nitric oxide synthase-2 in human optic nerve head astrocytes induced by elevated pressure in vitro, Arch. Ophthalmol. 119 (2001) 240 – 245. [12] H. Mi, B.A. Barres, Purification and characterization of astrocyte precursor cells in the developing rat optic nerve, J. Neurosci. 19 (1999) 1049 – 1061. [13] R.H. Miller, B.P. Williams, J. Cohen, M.C. Raff, A4: an antigenic marker of neural tube-derived cells, J. Neurocytol. 13 (1984) 329 – 338. [14] J.R. Naegele, C.J. Barnstable, A carbohydrate epitope defined by monoclonal antibody VC1.1 is found on N-CAM and other cell adhesion molecules, Brain Res. 559 (1991) 118 – 129. [15] A.H. Neufeld, Microglia in the optic nerve head and the region of parapapillary chorioretinal atrophy in glaucoma, Arch. Ophthalmol. 117 (1999) 1050 – 1056. [16] J.D. Pena, O.A. Agapova, B.T. Gabelt, L.A. Levin, M.J. Lucarelli, P.L. Kaufman, M.R. Hernandez, Increased elastin expression in astrocytes of the lamina cribrosa in response to elevated intraocular pressure, Investig. Ophthalmol. Vis. Sci. 42 (2001) 2303 – 2314. [17] G. Prasanna, R. Krishnamoorthy, A.F. Clark, R.J. Wordinger, T. Yorio, Human optic nerve head astrocytes as a target for endothelin-1, Investig. Ophthalmol. Vis. Sci. 43 (2002) 2704 – 2713. [18] C.S. Ricard, J.D. Pena, M.R. Hernandez, Differential expression of neural cell adhesion molecule isoforms in normal and glaucomatous human optic nerve heads, Mol. Brain Res. 74 (1999) 69 – 82. [19] C.S. Ricard, S. Kobayashi, J.D. Pena, M. Salvador-Silva, O.A. Agapova, M.R. Hernandez, Selective expression of neural cell adhesion molecule (NCAM)-180 in optic nerve head astrocytes exposed to elevated hydrostatic pressure in vitro, Mol. Brain Res. 81 (2000) 62 – 79. [20] M. Salvador-Silva, C.S. Ricard, O.A. Agapova, P. Yang, M.R. Hernandez, Expression of small heat shock proteins and intermediate filaments in the human optic nerve head astrocytes exposed to
76
P. Yang, M.R. Hernandez / Brain Research Protocols 12 (2003) 67–76
elevated hydrostatic pressure in vitro, J. Neurosci. Res. 66 (2001) 59 – 73. [21] G. Tezel, M.B. Wax, Increased production of tumor necrosis factoralpha by glial cells exposed to simulated ischemia or elevated hydrostatic pressure induces apoptosis in cocultured retinal ganglion cells, J. Neurosci. 20 (2000) 8693 – 8700. [22] M.B. Wax, G. Tezel, Neurobiology of glaucomatous optic neuropathy: diverse cellular events in neurodegeneration and neuroprotection, Mol. Neurobiol. 26 (2002) 45 – 55.
[23] R.J. Wordinger, R. Agarwal, M. Talati, J. Fuller, W. Lambert, A.F. Clark, Expression of bone morphogenetic proteins (BMP), BMP receptors, and BMP associated proteins in human trabecular meshwork and optic nerve head cells and tissues, Mol.Vis. 8 (2002) 241 – 250. [24] H. Ye, M.R. Hernandez, Heterogeneity of astrocytes in human optic nerve head, J. Comp. Neurol. 362 (1995) 441 – 452.