Bram Research, 89 (1975) 363-367
363
© Elsewer Scientific Pubhshmg Company, Amsterdam - Printed m The Netherlands
Immunofluorescence staining of astrocytes in vitro using antiserum to giial fibrillary acidic protein
DALE S ANTANITUS, BEN H. CHOI ANI) LOWELL W. LAPHAM Department of Pathology, Neuropathology Dtvlsion, Umverstty of Rochester Medtcal Center, Rochester, N. Y. 14642 (U.S.A.)
(Accepted February 18th, 1975)
Ghal fibrlllary acidic (GFA) protein has been demonstrated to be a major constituent of astrocytes. Utilizing immunofluorescence staining of GFA protein, features of normal fibrous astrocytes, reactive fibrous astrocytes, neoplastic astrocytes and differentiation of astrocytes during development have been elucidated 1-5. Until the present report, immunofluorescence staining of normal astrocytes has been applied successfully only to cryostat sections of frozen material. We report here the selective staining of astrocytes in primary tissue culture expiants of human fetal central nervous system (CNS) utilizing antiserum to human G F A protein in the redirect immunofluorescence technique of CoonsL Tissue cultures were prepared on glass coverslips as previously reported 6 from primitive mid-convexity (frontoparietal) cerebrum of 5 human fetuses (gestational ages 12-20 weeks) obtained as surgical pathology specimens from hysterotomies. After sufficient numbers of cells had become established in outgrowths from explants as determined by phase microscopy (between 8 and 39 days), coverslips were removed from Petri dishes and briefly rinsed in distilled water at room temperature. They were then quick-frozen on a cryostat chuck precooled to --20 °C. In order to ensure sufficient disruption of cell membranes to expose intracellular antigens, cover slips were brought to room temperature after ice crystals had formed. They were allowed to melt and then rapidly dried with forced stream room temperature air. Without further delay they were then returned to the cryostat and stored at --20 °C for 18-24 h, allowing subhmation of remaining moisture and facilitating concentration of antigen. The covershps were next fixed in precooled (--70 °C) acetone for 1 min and quickly air dried. This was followed by rinsing in phosphate buffered saline (PBS) for 2-3 mm at pH 7.2-7.3. The cells were then layered with 4 drops of rabb~t antiserum containing antibody to human G F A protein. Antiserum was diluted (1:50) with PBS Coverslips with antibody were incubated for 30 min at 37 °C in a high humidity chamber. (It was not necessary to control humidity precisely.) Excess antibody was removed by placing in a fresh PBS rinse for 4 mln. The coverslips were then layered
364
F~g 1 Phase contrast p h o t o m i c r o g r a p h ot m~gratmg neurons (allows) and aslmc~ies (41 ~, the outgro,,,,th zone of h u m a n fetal celebrum 9 day', m l ' m o (DIV) Gestational age ~as 14 weeks N,~tc a b u n d a n t polygonal cytoplasm, large pale nuclm and p l o m m e n t nucleoh of a,,liocytes Nt_,tll/ms are small, dark and possess long slender processes 300 F~g 2 Photomicrograph demon,~tratmg Sliver-posture processes of neurons (thin at rows) and Im gc~ astrocytlc nuclei (thick arro\~s) ~ t h no visible c3,toplasm in the outgrowth zone of htlrl],lll l~'l,l] cerebrum, 39 DIV Gestatlonal age, 14 ~,eeks Bodlan stain 480 Figs 3 and 4 [nlmtlllc*l]Ll¢*ic,,cencc of GI-A plo{eln In the cytoplasm and ploce~,',es ot asiit.~v~.lt'x Note delicate and condensed lib dlar paltci n q / a r e a s of intense fluolescence HtNnan fetal ccicln urn.
365 with 4 drops of fluorescein labeled goat antlrabbit gamma-globulin* (diluted 1:10) and again incubated for 30 mln at 37 °C in a high humidity chamber, followed by another 4-mln rinse in fresh PBS. Finally, coverslips were mounted in a solution containing 9 parts glycerol-1 part PBS and examined under a Zelss fluorescence microscope using an HBO 200 W/4 super pressure mercury burner light source, BG 38 and BG 12 exciter filters and a 500 nm barrier filter. Selected covershps were counterstained by first removing excess labeled antibody with a 2-mln rinse in PBS and placed in a fresh solution of 0.03 To methyl green in PBS for 2 min. These coverslips were again rinsed for 2 min in fresh PBS before mounting (With this counterstain nuclei fluoresced red.) Controls were processed by substituting PBS or normal rabbit serum for G F A protein antiserum or by substituting PBS for fluoresceln-labeled antibody. Further controls consisted of cultures explanted in an identical manner from human fetal leptomeninges. These were harvested at the same time as cultures of cerebrum and treated by the same experimental procedures. Similarly, cultures of human fibroblast line Wl 38 were exposed to the same lmmunofluorescent staining. By phase contrast microscopy the monolayer outgrowth immediately surroundlng the explant in the cultures of human fetal cerebrum revealed many small cells with small (6-9 #m), dark nuclei, scanty cytoplasm, and slender processes (Fig I). The processes of these cells were sllver-posmve in Bodian preparations (Fig 2), and the cells were interpreted as immature neurons. Immature neurons demonstrated no fluorescence or weak, non-specific fluorescence. With electron microscopy, the hallmark of this cell type was a unipolar process, containing a predominance of approximately 20-25-nm microtubules (Fig. 6) and sparse filaments. A second population, equally numerous, consisted of cells which were larger than the small cells, as observed by phase contrast microscopy. Nuclear diameters ranged between 15 and 20 #m. Cytoplasmic morphology was highly variable. Some possessed broad, sheet-like cytoplasm, others gave rise to processes, some quite long, with frequent branching. Intermediate forms between these extremes were recogmzed. Cells in this second population exhibited brilliant, green fluorescence (F~gs. 3 and 4). In cells with broad, sheet-like cytoplasm, fluorescence was often strongest in paranuclear regions or at the advancing cytoplasmic projections (Fig. 3) Cells with branching processes were usually more diffusely fluorescent, both in the region of the perikaryon and in the processes (Fig. 4). Cells which were fluorescent with antiserum to GFA could be correlated with cells in a similar location with large nuclei and variable cytoplasmic configuration in electron mlcrographs. As seen in Fig 5, these cells conslstently contained numerous sweeping bundles of approximately 6-9 nm filaments In silver preparations (Fig. 2), they were not impregnated. By correlation of the results of bright-field, phase, fluorescent and electron microscopic observations, this second cell population was identified as immature astrocytes. Control cultures, including cultures of fetal leptomenlnges and WI 38 fibroblasts, revealed either no fluorescence or low intensity non-specific fluorescence. * From Antibodies Incorporated, Daws, Calif.
Fig. 5. Electron micrograph of an astrocyte. Note abundant glial filaments (GF) measuring approximately 6-9 nm (arrows) in diameter within the cytoplasm. Human fetal cerebrum, 7 DIV. Gestational age, 18 weeks. × 37,500. Fig. 6. Electron micrograph of a neuron. Note scanty amount of cytoplasm and long slender neuritic process containing parallel arrays of neurotubules (arrows) measuring approximately 20-25 nm in diameter. Human fetal cerebrum, 11 DIV. Gestational age, 18 weeks. × 7,650.
367 The selective staining of astrocytes m cryostat sections by the indirect l m m u n o fluorescence techmque using antisera produced against the G F A protein has been established on the basis of studies by Blgnaml et al. 4, a n d Bignaml a n d Dahl ~'. The present report ~s a successful a p p h c a t m n of this ~mmunofluorescent technique for the d e m o n s t r a t i o n of astrocytes in tissue cultures of the CNS. U t i h z i n g this method it is now possible to have greater confidence in the identification of astrocytes m rttro. A l t h o u g h it is n o t possible to state that every astrocyte in our cultures fluoresced, structures resembling i m m a t u r e n e u r o n s clearly did not fluoresce Since leptomeningeal cells a n d fibroblasts derived from l e p t o m e n m g e s or from blood vessels often create a p r o b l e m in cell identification in tissue cultures, it was particularly helpful to find that our cultures of leptomeninges and WI 38 fibroblasts exposed to the i m m u n o fluorescent techmque exhibited either no fluorescence or slight, non-specific fluorescence. These observations are in keeping with the selectiwty of this i m m u n o f l u o r e s c e n t method for astrocytes m the studies of Blgnaml, Dahl a n d their collaborators I 5,s,9 In s u m m a r y , identification of astrocytes in vitro is greatly facihtated by the use of l m m u n o f l u o r e s c e n c e staining for G F A protein. Th~s should prove to be a powerful tool in tissue culture work on the CNS, where identification of cell types is a m a j o r p r o b l e m a n d positive criteria at the light microscope level have been limited. We t h a n k Dr. A m l c o Bignaml for supplying us with a n t i s e r u m to G F A protein. We also wish to t h a n k Miss Rose Rosh a n d Miss M a r t h a Davis for techmcal assistance, a n d Mrs. L m d a C r a n d a l l for typing the m a n u s c r i p t Supported in part by U.S. Public Health Service G r a n t No. 5 R01 H D 07078 from N 1 C H D .
1 BIGNAMI, A., AND DAHL, D, Dlfferenhation of astrocytes m the cerebellar cortex and the pyramidal tracts of the newborn rat An lmmunofluorescence study with antibodies to a protein specific to astrocytes, Brain Research, 49 (1973) 393-402 2 BIGNAMI, A., AND DAHL, D, Astrocyte-specific protein and neuroghal differentiation An Immunofluorescence study with antibodies to the ghal fibrillary acidic protein, J. comp Neurol, 153 (1974) 27-38 3 BIGNAML A , AND DAHL, D, Astrocyte-specLfic protein and radial gha in the cerebral cortex of newborn rat, Nature (Lond), 252 (1974) 55-56 4 BIGNAMI, A, E~G, L F , DAHL, D, A~D UYEDA,C T , Locahzatlon of the ghal fibrillary acidic protein in a~trocytes by immunofluorescence, Btam Research, 43 (1972) 429-435. 5 BISSELL,M G , RUBINSTEIN, L J, BIG~AMI,A , AND HERMAN, M M, Characteristics of the rat C-6 glioma maintained in organ culture systems Production of ghal fihrillary acidic protein m the absence of ghofibrdlogenesis, Brain Research, 82 (1974) 77 89 6 CHOL B H, AND LAPHAM, L W, Autoradmgraphic studies of migrating neurons and astrocytes of huinan fetal cerebral cortex m vttro, Eap molec Pathol, 21 (1974) 204-217 7 CooNs, A H, AND KAPLAN, M. H., Locahzatlon of antigen in tissue cells. 11. Improvements in a method for the detection of antigen b~ means of fluorescent antibody, J exp Med, 91 (1950) 1-13 8 DAHL, D, AND BIGNAMI, A, Ghal fibrillary acidic protein from normal human brain Purification and properties, Btam Resea;ch, 57 (1973) 343-360 9 ENC., L F , VANDERHAEGHEN, J J., BIGNAMI, A, AND GERSTL, B , An acidic protein isolated from fibrous astrocytes, Brain Research, 28 (1971) 351-354,
363
© Elsewer Scientific Pubhshmg Company, Amsterdam - Printed m The Netherlands
Immunofluorescence staining of astrocytes in vitro using antiserum to giial fibrillary acidic protein
DALE S ANTANITUS, BEN H. CHOI ANI) LOWELL W. LAPHAM Department of Pathology, Neuropathology Dtvlsion, Umverstty of Rochester Medtcal Center, Rochester, N. Y. 14642 (U.S.A.)
(Accepted February 18th, 1975)
Ghal fibrlllary acidic (GFA) protein has been demonstrated to be a major constituent of astrocytes. Utilizing immunofluorescence staining of GFA protein, features of normal fibrous astrocytes, reactive fibrous astrocytes, neoplastic astrocytes and differentiation of astrocytes during development have been elucidated 1-5. Until the present report, immunofluorescence staining of normal astrocytes has been applied successfully only to cryostat sections of frozen material. We report here the selective staining of astrocytes in primary tissue culture expiants of human fetal central nervous system (CNS) utilizing antiserum to human G F A protein in the redirect immunofluorescence technique of CoonsL Tissue cultures were prepared on glass coverslips as previously reported 6 from primitive mid-convexity (frontoparietal) cerebrum of 5 human fetuses (gestational ages 12-20 weeks) obtained as surgical pathology specimens from hysterotomies. After sufficient numbers of cells had become established in outgrowths from explants as determined by phase microscopy (between 8 and 39 days), coverslips were removed from Petri dishes and briefly rinsed in distilled water at room temperature. They were then quick-frozen on a cryostat chuck precooled to --20 °C. In order to ensure sufficient disruption of cell membranes to expose intracellular antigens, cover slips were brought to room temperature after ice crystals had formed. They were allowed to melt and then rapidly dried with forced stream room temperature air. Without further delay they were then returned to the cryostat and stored at --20 °C for 18-24 h, allowing subhmation of remaining moisture and facilitating concentration of antigen. The covershps were next fixed in precooled (--70 °C) acetone for 1 min and quickly air dried. This was followed by rinsing in phosphate buffered saline (PBS) for 2-3 mm at pH 7.2-7.3. The cells were then layered with 4 drops of rabb~t antiserum containing antibody to human G F A protein. Antiserum was diluted (1:50) with PBS Coverslips with antibody were incubated for 30 min at 37 °C in a high humidity chamber. (It was not necessary to control humidity precisely.) Excess antibody was removed by placing in a fresh PBS rinse for 4 mln. The coverslips were then layered
364
F~g 1 Phase contrast p h o t o m i c r o g r a p h ot m~gratmg neurons (allows) and aslmc~ies (41 ~, the outgro,,,,th zone of h u m a n fetal celebrum 9 day', m l ' m o (DIV) Gestational age ~as 14 weeks N,~tc a b u n d a n t polygonal cytoplasm, large pale nuclm and p l o m m e n t nucleoh of a,,liocytes Nt_,tll/ms are small, dark and possess long slender processes 300 F~g 2 Photomicrograph demon,~tratmg Sliver-posture processes of neurons (thin at rows) and Im gc~ astrocytlc nuclei (thick arro\~s) ~ t h no visible c3,toplasm in the outgrowth zone of htlrl],lll l~'l,l] cerebrum, 39 DIV Gestatlonal age, 14 ~,eeks Bodlan stain 480 Figs 3 and 4 [nlmtlllc*l]Ll¢*ic,,cencc of GI-A plo{eln In the cytoplasm and ploce~,',es ot asiit.~v~.lt'x Note delicate and condensed lib dlar paltci n q / a r e a s of intense fluolescence HtNnan fetal ccicln urn.
365 with 4 drops of fluorescein labeled goat antlrabbit gamma-globulin* (diluted 1:10) and again incubated for 30 mln at 37 °C in a high humidity chamber, followed by another 4-mln rinse in fresh PBS. Finally, coverslips were mounted in a solution containing 9 parts glycerol-1 part PBS and examined under a Zelss fluorescence microscope using an HBO 200 W/4 super pressure mercury burner light source, BG 38 and BG 12 exciter filters and a 500 nm barrier filter. Selected covershps were counterstained by first removing excess labeled antibody with a 2-mln rinse in PBS and placed in a fresh solution of 0.03 To methyl green in PBS for 2 min. These coverslips were again rinsed for 2 min in fresh PBS before mounting (With this counterstain nuclei fluoresced red.) Controls were processed by substituting PBS or normal rabbit serum for G F A protein antiserum or by substituting PBS for fluoresceln-labeled antibody. Further controls consisted of cultures explanted in an identical manner from human fetal leptomeninges. These were harvested at the same time as cultures of cerebrum and treated by the same experimental procedures. Similarly, cultures of human fibroblast line Wl 38 were exposed to the same lmmunofluorescent staining. By phase contrast microscopy the monolayer outgrowth immediately surroundlng the explant in the cultures of human fetal cerebrum revealed many small cells with small (6-9 #m), dark nuclei, scanty cytoplasm, and slender processes (Fig I). The processes of these cells were sllver-posmve in Bodian preparations (Fig 2), and the cells were interpreted as immature neurons. Immature neurons demonstrated no fluorescence or weak, non-specific fluorescence. With electron microscopy, the hallmark of this cell type was a unipolar process, containing a predominance of approximately 20-25-nm microtubules (Fig. 6) and sparse filaments. A second population, equally numerous, consisted of cells which were larger than the small cells, as observed by phase contrast microscopy. Nuclear diameters ranged between 15 and 20 #m. Cytoplasmic morphology was highly variable. Some possessed broad, sheet-like cytoplasm, others gave rise to processes, some quite long, with frequent branching. Intermediate forms between these extremes were recogmzed. Cells in this second population exhibited brilliant, green fluorescence (F~gs. 3 and 4). In cells with broad, sheet-like cytoplasm, fluorescence was often strongest in paranuclear regions or at the advancing cytoplasmic projections (Fig. 3) Cells with branching processes were usually more diffusely fluorescent, both in the region of the perikaryon and in the processes (Fig. 4). Cells which were fluorescent with antiserum to GFA could be correlated with cells in a similar location with large nuclei and variable cytoplasmic configuration in electron mlcrographs. As seen in Fig 5, these cells conslstently contained numerous sweeping bundles of approximately 6-9 nm filaments In silver preparations (Fig. 2), they were not impregnated. By correlation of the results of bright-field, phase, fluorescent and electron microscopic observations, this second cell population was identified as immature astrocytes. Control cultures, including cultures of fetal leptomenlnges and WI 38 fibroblasts, revealed either no fluorescence or low intensity non-specific fluorescence. * From Antibodies Incorporated, Daws, Calif.
Fig. 5. Electron micrograph of an astrocyte. Note abundant glial filaments (GF) measuring approximately 6-9 nm (arrows) in diameter within the cytoplasm. Human fetal cerebrum, 7 DIV. Gestational age, 18 weeks. × 37,500. Fig. 6. Electron micrograph of a neuron. Note scanty amount of cytoplasm and long slender neuritic process containing parallel arrays of neurotubules (arrows) measuring approximately 20-25 nm in diameter. Human fetal cerebrum, 11 DIV. Gestational age, 18 weeks. × 7,650.
367 The selective staining of astrocytes m cryostat sections by the indirect l m m u n o fluorescence techmque using antisera produced against the G F A protein has been established on the basis of studies by Blgnaml et al. 4, a n d Bignaml a n d Dahl ~'. The present report ~s a successful a p p h c a t m n of this ~mmunofluorescent technique for the d e m o n s t r a t i o n of astrocytes in tissue cultures of the CNS. U t i h z i n g this method it is now possible to have greater confidence in the identification of astrocytes m rttro. A l t h o u g h it is n o t possible to state that every astrocyte in our cultures fluoresced, structures resembling i m m a t u r e n e u r o n s clearly did not fluoresce Since leptomeningeal cells a n d fibroblasts derived from l e p t o m e n m g e s or from blood vessels often create a p r o b l e m in cell identification in tissue cultures, it was particularly helpful to find that our cultures of leptomeninges and WI 38 fibroblasts exposed to the i m m u n o fluorescent techmque exhibited either no fluorescence or slight, non-specific fluorescence. These observations are in keeping with the selectiwty of this i m m u n o f l u o r e s c e n t method for astrocytes m the studies of Blgnaml, Dahl a n d their collaborators I 5,s,9 In s u m m a r y , identification of astrocytes in vitro is greatly facihtated by the use of l m m u n o f l u o r e s c e n c e staining for G F A protein. Th~s should prove to be a powerful tool in tissue culture work on the CNS, where identification of cell types is a m a j o r p r o b l e m a n d positive criteria at the light microscope level have been limited. We t h a n k Dr. A m l c o Bignaml for supplying us with a n t i s e r u m to G F A protein. We also wish to t h a n k Miss Rose Rosh a n d Miss M a r t h a Davis for techmcal assistance, a n d Mrs. L m d a C r a n d a l l for typing the m a n u s c r i p t Supported in part by U.S. Public Health Service G r a n t No. 5 R01 H D 07078 from N 1 C H D .
1 BIGNAMI, A., AND DAHL, D, Dlfferenhation of astrocytes m the cerebellar cortex and the pyramidal tracts of the newborn rat An lmmunofluorescence study with antibodies to a protein specific to astrocytes, Brain Research, 49 (1973) 393-402 2 BIGNAMI, A., AND DAHL, D, Astrocyte-specific protein and neuroghal differentiation An Immunofluorescence study with antibodies to the ghal fibrillary acidic protein, J. comp Neurol, 153 (1974) 27-38 3 BIGNAML A , AND DAHL, D, Astrocyte-specLfic protein and radial gha in the cerebral cortex of newborn rat, Nature (Lond), 252 (1974) 55-56 4 BIGNAMI, A, E~G, L F , DAHL, D, A~D UYEDA,C T , Locahzatlon of the ghal fibrillary acidic protein in a~trocytes by immunofluorescence, Btam Research, 43 (1972) 429-435. 5 BISSELL,M G , RUBINSTEIN, L J, BIG~AMI,A , AND HERMAN, M M, Characteristics of the rat C-6 glioma maintained in organ culture systems Production of ghal fihrillary acidic protein m the absence of ghofibrdlogenesis, Brain Research, 82 (1974) 77 89 6 CHOL B H, AND LAPHAM, L W, Autoradmgraphic studies of migrating neurons and astrocytes of huinan fetal cerebral cortex m vttro, Eap molec Pathol, 21 (1974) 204-217 7 CooNs, A H, AND KAPLAN, M. H., Locahzatlon of antigen in tissue cells. 11. Improvements in a method for the detection of antigen b~ means of fluorescent antibody, J exp Med, 91 (1950) 1-13 8 DAHL, D, AND BIGNAMI, A, Ghal fibrillary acidic protein from normal human brain Purification and properties, Btam Resea;ch, 57 (1973) 343-360 9 ENC., L F , VANDERHAEGHEN, J J., BIGNAMI, A, AND GERSTL, B , An acidic protein isolated from fibrous astrocytes, Brain Research, 28 (1971) 351-354,
