simultaneous immunofluorescence and autoradiography: a useful technique for investigating...

Developmental Brain Research, 36 (1987) 1-11 1 Elsevier

BRD 50606

Research Reports

Simultaneous immunofluorescence and autoradiography" a useful technique for investigating neurotransmitter uptake by neurons

and glia in primary central nervous system culture

Richard Reynolds and Norbert Herschkowitz Department of Pediatrics, University of Berne, Inselspital, Berne (Switzerland)

(Accepted 24 February 1987)

Key words: Immunocytochemistry; Autoradiography; Oligodendrocyte; Astrocyte; Neuron; V-[3H]Aminobutyric acid; D-[3H]Aspartate; Tissue culture

Previous studies on the localization of radiolabelled neurotransmitters in cultured cells of neural origin have relied on the comparison of cell morphology, as determined by immunocytochemistry, with the patterns of labelling on autoradiograms. We present here a method combining simultaneously autoradiography, following the uptake of tritium-labelled amino acid transmitters, with indirect immunofluorescence using antibodies against both surface and intracellular antigens. Using a fixative containing only a low concentration of glutaraldehyde (4% paraformaldehyde, 0.1% glutaraldehyde), a similar retention of V-[3H]aminobutyric acid (GABA) and D-[3H]aspartate was achieved as with the higher concentrations commonly used, with the advantage that the autofluorescence asso- ciated with glutaraldehyde fixed tissue was eliminated, and the immunoreactivity of the antigens to be localized was not destroyed. Using this method GABA and D-aspartate accumulating cells, in dissociated mouse central nervous system (CNS) cultures, could be reliably identified as oligodendrocytes, and some multiprocessed astrocytes, by anti-galactocerebroside (GC) and anti-glial fibrillary acidic protein (GFAP) immunofluorescence respectively. GABA-accumulating neuron-specific enolase (NSE) positive neurons could be clearly identified but no D-aspartate accumulating neurons were found. This technique should have a wide application in the investigation of whether selective transport mechanisms coexist with antigens characteristic of a certain cell type or sub-type.

INTRODUCTION

The inactivation of amino acid and monoamine neurotransmitters in the central nervous system

(CNS), with the exception of acetylcholine, following their release by neurons, is generally thought to occur by uptake via specific high-affinity transport

systems into both neurons and glia, rather than by ex-

tracellular enzymatic degradation. Evidence for the existence of such transport systems has come from autoradiographical localization after uptake of the radiolabelled transmitter in vivo 13, in vitro 12, and in culture 1°'16, and on quantitative in vitro estimation of

uptake into isolated cell preparations 6'11. Recently

primary dissociated CNS cultures have become a useful tool and offer significant advantages, for the investigation of neurotransmitter uptake and metab-

olism by different neural cell types during devel-

opment. Autoradiographical studies using primary

cultures have demonstrated the possible importance

of CNS glia, in addition to neurons, in the inactivation of amino acid transmitters 7'2°'21'26 and mono-

amines 14,2s. Due to the heterogeneous nature of pri-

mary CNS cultures and the ambiguity of the identifi-

cation of cultured cells, at the light microscopic level, by morphology alone, such autoradiographical stud-

ies are usually performed alongside immunocyto-

chemical characterization of cell populations using cell specific antigens 2°'21'25'26.

These previous studies have relied on the comparison of the morphology and distribution of the cell

types, as recognized by immunocytochemistry, with the pattern of labelling of the autoradiograms 4,20,21,25,26. However , different cell types that

express similar morphologies (e.g. oligodendroglia and process-bearing astroglia) may still be difficult to

Correspondence: R. Reynolds. Present address: Department of Biochemistry, Imperial College, London, SW7 2AZ, U.K.

0165-3806/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

distinguish using these methods. Definitive identification of individual cells that become labelled with tritiated neurotransmitters requires the simultaneous combination of autoradiography with immunocytochemical staining. Several studies have recently tried this approach and have found that it is not possible to simply combine the two techniques 2,18. Chemical fixatives must be used that allow the retention of suffi- cient radiolabel, do not cause autofluorescence, and do not destroy the immunoreactivity of the marker antigens. Several studies combining 7-[3H]aminobu - tyric acid ([3H]GABA)1,2,22 or [3H]serotonin15 auto-

radiography with immunofluorescence have, however, successfully localized the transmitter to particular cell types, but the methods used have not been ap- plied, or are not applicable, to more than one antigen or transmitter. Such a technique has also successfully been used to localize/3-adrenergic receptors on cultured cells identified with cell type-specific immunofluorescence 3. The combination of [3H]GABA autoradiography with peroxidase-antiperoxidase immu- nohistochemistry has also been used in several recent studies 8'18'19, which avoids the problem of autofluo-

rescence caused by glutaraldehyde containing fixatives, but has the disadvantage that the peroxidase reaction product is too easily obscured by the overlying silver grains. Therefore this method allows only very low silver grain densities and necessitates the counting of silver grains to establish the presence of double-labelled cells 19.

We present here a technique that makes it possible to visualize the uptake of the amino acids [3H]GABA and D-[3H]aspartate by autoradiography in neurons, astrocytes and oligodendrocytes, in primary CNS mixed cell cultures, simultaneously identified by indirect immunofluorescence against both surface and intracellular antigens. This technique has a wide application in studies of the coexistence of selective uptake mechanisms with cell type-specific antigens and can be used to investigate subpopulations of one cell type and the developmental expression of such uptake mechanisms (see companion paper ref. 23).

MATERIALS AND METHODS

Cell cultures Primary mixed cell cultures were prepared from

forebrains of neonatal or embryonic day 15 Swiss al-

bino mice as described previ0usly 2°'21, and grown on poly-L-lysine-coated glass coverslips in 35-mm Fal- con Multiwell dishes (1 x 105 viable cells/cm2). Cul- ture medium consisted of Dulbecco's Modified Eagles Medium (DMEM) containing 10% fetal calf serum (FCS), 30 mM glucose, 7.4 mM L-glutamine and 100 U/ml penicillin, and the medium changed after 4 days and thereafter every 3 days. Neuron-enriched cultures were prepared from forebrains of embryonic day 15 mice as described above and after 24 h growth in DMEM plus 10% FCS were transferred to

a serum-free culture medium consisting of DMEM/ Hams F12, 1:1 (vol/vol), containing (final concentrations) 10 mM HEPES buffer, 12 mM NaHCO3, 30 mM glucose, 4.2 mM L-glutamine, 10 ktg/ml insulin, 50/~g/ml transferrin, 100/~M putrescine, 30 nM sele- nium, 20 nM progesterone and 15 nM triiodothyro- nine. Cultures were grown until 14 days in vitro (DIV) in this medium, with complete medium changes every 3 days.

Antisera The rabbit anti-galactocerebroside (anti-GC) se-

rum was prepared and tested as described previously 21, and was used for indirect immunofluorescence of living cells at a dilution of 1:10. Rabbit anti-glial fibrillary acidic protein (anti-GFAP) and rabbit anti- neuron specific enolase (anti-NSE) were purchased from DAKO and Polysciences respectively, and used at dilutions of 1:50 and 1:200 respectively. Non-im- mune rabbit serum was used at corresponding dilutions and gave no positive reaction in all cases.

Surface immunofluorescence and autoradiography Cultures were rinsed twice in Hank's balanced salt

solution (HBSS) and incubated for 20 min (20 °C) with primary antiserum diluted in HBSS containing 5% normal goat serum. The cultures were then washed 4 times in HBSS over 10 min and incubated for 20 min (20 °C) with tetramethylrhodamine iso- thiocyanate-conjugated goat anti-rabbit IgG (TRITC-GAR IgG; Cappel Labs., IgG fraction), diluted 1:150 in HBSS. After two rinses in HBSS and two rinses at 37 °C in Tris-buffered Krebs salt solution 21 (TBSS, pH 7.4) the cultures were preincubated for 5 min at 37 °C. [3H]GABA (6 x 10 -8 M, 30.8 Ci/mmol, NEN) and D-[3H]aspartate (1.4 x 10 -7 M, 14.0 Ci/mmol, NEN) were then added at 2 ktCi/ml

and incubation continued for 10 min. Incubation was terminated by 4 washes with sodium-free TBSS (20 °C), containing 1 mM unlabelled amino acid to minimise both specific and non-specific binding to membrane receptors. Cultures were then fixed with 4% paraformaldehyde, 0.1% glutaraldehyde (EM grade, Fluka), in 0.1 M S6rensens sodium phosphate buffer (pH 7.2) for 20 min, followed by 4 washes in phosphate buffered saline (PBS).

Intracellular immunofluorescence and autoradiography

For localization of intracellular antigens combined with [3H]amino acid autoradiography a method similar to that described recently by Kimelberg and Katz 15 was used. Cultures were first incubated with [3H]GABA or D-[3H]aspartate for 1-20 min as

described above. (a) GFAP. For the localization of GFAP the cul-

tures were fixed with 4% paraformaldehyde 0.1% glutaraldehyde for 20 min, following 4 washes with sodium-free TBSS. After two washes (2 min each) with PBS the cultures were permeabilized by a 10- min incubation in methanol at -20 °C, followed by 4 washes (2 min each) with PBS (20 °C).

(b) NSE. For the localization of NSE cultures were fixed for 30 min with 4% paraformaldehyde 0.1% glutaraldehyde 0.2% picric acid in 0.1 M sodium phosphate buffer (pH 7.2) 24, washed twice with

PBS and the cultures permeabilized by a 5-min incubation in 0.25% Triton X-100 in PBS.

After a further 4 washes with PBS all cultures were incubated for 30 min with 10% normal goat serum in PBS, washed 4 times with PBS, and incubated 30 min with the primary antiserum, diluted in PBS containing 5% normal goat serum. Control slides were incubated with normal rabbit serum. After 4 washes with PBS cultures were left for 30 min in PBS, which was changed to fresh solution every 10 min. Incubation for 30 min with TRITC-GAR IgG in PBS was followed by washing, as after the primary antiserum.

Autoradiography Cultures were rinsed once in distilled water and,

either air dried, or dehydrated in an ethanol series. The coverslips were covered with Ilford K2 emulsion (1:1 with water) and exposed for 7 days (NSE), 10 days (cell surface antigens), or 12 days (other intra-

cellular antigens) at 4 °C. The coverslips were devel- oped as described previously 2°, and a second glass coverslip mounted on the first.

Microscopy Slides were examined with a Leitz Orthoplan mi-

croscope using bright field and fluorescence optics (filter block N2) and × 25 and × 40 oil immersion ob- jectives. Photographs were taken with Ilford HP5 400ASA black and white films.

RESULTS

The cellular composition of the dissociated mixed cell cultures used in this study, prepared from both fetal and neonatal mouse brain, has been described previously in detail 2°'21.

Effects of fixation on immunofluorescence Tetramethylrhodamine (TRITC)-labelled second

antibodies were used throughout this study as it was found in preliminary experiments that fading of the fluorescence from fluorescein labelled antibodies was so extensive after dehydration, covering with emulsion, exposure and developing, that it was no longer possible to obtain clear photographs. Al- though such extensive fading does not occur with TRITC, or Texas Red 22, there was a clear reduction

in the fluorescence intensity after the whole autoradiographic procedure.

After fixation with paraformaldehyde containing 0.1% glutaraldehyde, following incubation of living cells with both primary and secondary antibodies prior to the fixation, the resulting immunofluorescence was a little less intense than in the absence of glutaraldehyde. The intense autofluorescence that occurs after fixation with higher concentrations of glutaraldehyde 2 was reduced to very low levels by re- ducing the concentration to 0.1%. For the immuno- fluorescent labelling of surface antigens on living cells glutaraldehyde obviously does not affect the immunoreactivity of the antigens, as long as both the primary and secondary antibody incubations are car- ried out prior to fixation.

For the immunofluorescence labelling of intracellular antigens the presence of 0.1% glutaraldehyde did not destroy the immunoreactivity of the antigens localized in the present study, although there was a

clear reduction in the fluorescence intensity with

GFAP. In contrast with NSE localization the inclu- sion of 0.1% glutaraldehyde in the fixative increased the intensity of the specific fluorescence. The inclu-

sion of glutaraldehyde also caused an increase in the

amount of non-specific fluorescence, which was not due to autofluorescence. However , this could be greatly reduced by including an incubation with 10% normal goat serum prior to the primary antibody and

including 5% normal goat serum with the primary an-

tibody. The method used to permeabilize the cells to allow access of the antibodies to the intracellular antigens was also important. Permeabilization using the

detergent Triton X-100 produced a high non-specific background. This was greatly reduced when, either

acetone (50% 1 rain, 100% 1 min, and 50% 1 min; -20 °C), or methanol (10 min, -20 °C), were used for

this step. Use of acetic acid/ethanol (5:95, 10 min, -20 °C) resulted in a lower intensity of the specific immunofluorescence with all antigens when com-

pared with methanol. In further experiments a 10-

min incubation with methanol at -20 °C was judged to give the brightest specific immunofluorescence and the lowest background, and therefore this method was used in all further experiments for

G F A P localization. For NSE it was necessary to use Triton X-100 for the permeabilization step because

methanol produced a large reduction in the specific immunofluorescence.

The use of benzoquinone as primary fixative, as in a previous study 2, resulted in very poor specific fluorescence with all antigens.

Effects of fixation on autoradiography Substitution of 2% glutaraldehyde, the fixative of

choice for light microscopic autoradiography, with

4% paraformaldehyde, the fixative generally used in

indirect immunofluorescence, resulted in consider-

able reductions in the retention of both [3H]GABA and D-[3H]aspartate (Table I), as has been noted previously 2'2°. Addition of as little as 0.1% glutaral-

dehyde to the paraformaldehyde restored the retention of [3H]GABA to 42%, a value higher than that

for 2% glutaraldehyde alone (Table I). In contrast,

the retention of D-[3H]aspartate was reduced in the presence of 0.1% glutaraldehyde in comparison with 2%. These differences are also clearly apparent on autoradiograms (Fig. 1). After 10 min incubation

TABLE I

Degree of retention of [3H]GABA and o-[3H]aspartate by various fixatives

Cultures prepared from neonatal mouse brain, grown for 16 DIV, were incubated with [3H]amino acid for 10 min as described in the Materials and Methods section. The tissue on the glass coverslips was solubilized overnight in 0.5 N NaOH, after fixation for 20 rain. All fixatives were prepared fresh in 0.1 M sodium phosphate buffer (pH 7.2). Values are means + S.E.M. from 3 independent experiments (9 cultures). The 100% retention represents the 3H-content of the cultures before fixation. Control values were: GABA 35614 _+ 3303 dpm/min/mg protein (3.24 + 0.30 pmol/min/mg protein); D-aspartate 118342 _+ 12156 dpm/min/mg protein (10.76 + 1.11 pmol/min/mg protein).

Fixative [3H]GABA D-[3H]Aspartate

(a) 2% Glutaraldehyde (b) 4% Paraformaldehyde (c) 4% Paraformaldehyde

0.1% glutaraldehyde (d) 4% Paraformaldehyde,

0.1% glutaraldehyde, 0.2% picric acid

35.4__+ 1.5% 33.0 + 2.8% 12.7 + 0.6% 8.9 + 1.3% 42.3 __+ 0.1% 20.9 + 2.7%

35.5 + 2.2% 18.6 + 2.3%

with [3H]GABA or D-[3H]aspartate, and fixation

with 2% glutaraldehyde, the small process-bearing cells (predominantly oligodendrocytes) were nearly

all heavily labelled over both cell body and processes (Fig. la, c). Whereas fixation with 0.1% glutaral-

dehyde and 4% paraformaldehyde resulted in very little difference in the intensity of the [3H]GABA au- toradiogram (Fig. lb), the labelling intensity of both

process-bearing cells and flat epithelioid astrocytes with D-[3H]aspartate was greatly reduced, with the

labelling of the process-bearing cells occurring only over the cell bodies (Fig. ld). It is pertinent to men- tion here that if cells have only accumulated a small amount of radiolabel then after the loss of label that

occurs during fixation they may appear totally unla-

belled on autoradiograms. Thus the level of detecta-

bility may be a limitation with this method. The fixative used for NSE localization (4% paraformaldehyde, 0.1% glutaraldehyde, 0.2% picric acid) produced a similar degree of retention to that used for

G F A P (Table I).

Combined surface immunofluorescence and autoradiography

After 10 min incubation with [3H]GABA or D- [3H]aspartate the small process-bearing cells were

Fig. 1. Effect of reduced glutaraldehyde concentration on [3H]GABA and D-[3H]aspartate autoradiography. At 11 DIV neonatal mixed glial cultures were incubated for 10 min with [3H]GABA (a,b) or D-[3H]aspartate (e,d) and fixed with, either 2% glutaraldehyde (a,c), or 4% paraformaldehyde 0.1% glutaraldehyde (b,d). The labelling intensity of the process bearing cells (predominantly oligodendrocytes) and underlying astrocytes, with [3H]GABA, was not affected by the different fixatives, whereas the labelling intensity of both cell types with D-[3H]aspartate was greatly reduced (d) after fixation with the lower glutaraldehyde concentration. Fields are representative of the general results obtained. Exposure time 10 days. Bar = 30/~m.

labelled intensely enough for the cell bodies to be

densely covered with silver grains (Fig. 2b, d). A

large number of these cells were also labelled with

anti-GC antibodies, identifying them positively as

oligodendrocytes 21 (Fig. 2a, c). It was possible to

clearly see the brilliant membrane staining of the pro-

cesses, which were not as heavily labelled with silver

grains as the cell bodies, over which the fluorescence

was obscured by the silver grains.

To investigate the effect of the surface antigen

staining procedure on the uptake of the amino acids,

cultures were incubated with both the primary and

secondary antibodies followed by [3H]GABA or D-

[3H]aspartate and then solubilized (Table II). Prior

incubation with primary and secondary antibodies

had no significant effect on the ability of the cultures

to accumulate either amino acid. The use of normal

rabbit serum or ant i -GC serum as the primary anti-

body also showed no difference, indicating that the

TABLE II

Effect of surface antigen staining procedure on the uptake and retention of f H]GABA and o-[3H]aspartate

Cultures prepared from neonatal mouse brain, grown for 11 DIV, were treated as above before being incubated with [3H]GABA or D-[3H]aspartate, or in C treated with antiserum, incubated with radiolabel, and then fixed (20 min) with 4% paraformaldehyde 0.1% glutaraldehyde. Values are means + S.E.M. from 3 independent experiments (9 cultures). Con- trols (100% retention) represent values from cultures not incubated with antiserum but only with [3H]amino acid. Control values for GABA were 29748 + 2483 dpm/min/mg protein and for D-aspartate 131236 + 3706 dpm/min/mg protein.

Procedure [3H]GABA D-[3H]Aspartate (% of control) (% of control)

(a) Incubation with 10% 89.8 _+ 4.1 101.8 + 3.2 normal rabbit serum

(b) Incubation with 10% 87.5 + 4.2 98.6 + 1.8 rabbit anti-GC serum

(c) Incubation with 10% 47.3 _+ 2.9 29.0 _+ 2.0 anti-GC serum then fixation

Fig. 3. Simultaneous [3H]GABA and D-[3H]aspartate autoradiography and anti-GFAP immunofluorescence, a,b: 21-DIV culture, prepared from embryonic mouse brain, incubated 10 min with [3H]GABA. c,d: l l -DIV culture, prepared from neonatal mouse brain, incubated 10 min with D-[3H]aspartate. Two process-bearing glial cells are seen that are labelled over both cell body and processes after incubation with, either GABA (b), or D-aspartate (d), one of which has GFAP-positive processes (arrow) and the other not (arrowhead). The underlying epithelioid astrocytes, which are also GFAP-positive, are only sparsely covered with silver grains. Bar = 16 ktm (a,b) and 20¢tm (c,d). Exposure time 12 days.

b inding of the p r imary an t ibody to the m e m b r a n e lip-

id G C did no t in te r fe re with the func t ion o f the

G A B A and g lu t ama te t ranspor t pro te ins .

Combined intracellular immunofluorescence and autoradiography

Fol lowing the who le p r o c e d u r e of in t race l lu lar im-

Fig. 2. Simultaneous [3H]GABA and D-[3H]aspartate autoradiography and anti-GC immunofluorescence. Two oligodendrocytes, in mixed glial cultures 16 DIV (a,b) and 22 DIV (c,d), whose processes are clearly labelled with antibodies against the surface marker GC (a,c), are also heavily labelled with silver grains over the cell bodies after 10 min incubation with [3H]GABA (b) and D-[3H]aspar - tate (d). Exposure time 10 days. Bar = 16pm.

munofluorescence staining, the labell ing of the cells

with silver grains, after 10 min incubat ion with ei ther

[3H]amino acid, was greatly diminished when com-

pared with controls. The autoradiographic labelling

pat tern, although less intense than after the surface

antigen staining procedure , revealed similar results,

with many small process bearing cells still heavily

labelled over cell body and some lightly label led over

the processes. Less than 3% of these heavily label led

process-bear ing cells present in serum-containing

cultures were found to be label led with a n t i - G F A P

ant ibodies (both G A B A and D-aspartate-accumulat-

ing cells; Fig. 3), therefore identifying them as astro-

cytes. Again these double- label led cells could be

most clearly identif ied by their brightly f luorescent

processes that were not so densely covered with sil-

ver grains (Fig. 3b, d). When incubat ion t imes longer

than 10 min, or exposure t imes longer than 12 days,

Fig. 4. Simultaneous [3H]GABA and D-[3H]aspartate autoradiography and anti-NSE immunofluorescence of 14-DIV neuron-enriched cultures, a,b: two NSE-positive neurons are shown, one of which has accumulated GABA (arrowhead) after a 5-min incubation, while the other (arrow) is completely unlabelled. The underlying astrocytes are also completely unlabelled, c,d: in contrast, the underlying astrocytes are moderately labelled with D-aspartate after a 3-min incubation, whereas the NSE-positive neurons are completely free of silver grains. Bar = 20 #m. Exposure time 7 days.

were used, the covering of silver grains obscured the fluorescence to an extent that made the identification of double labelled cells difficult or unreliable. Only a slightly greater labelling intensity of flat GFAP-positive cells was obtained with increased incubation times (up to 60 min) and these cells never became more than lightly labelled in serum-containing cultures. Many neurons in the neuron enriched cultures became heavily labelled with [3H]GABA and it was necessary to use shorter incubation times of 1-5 min, and exposure times of 5 -7 days in order to be able to identify NSE-positive GABA-accumulating neurons. Fig. 4a, b shows both GABA-accumulating and non-GABA-accumulating NSE-positive neurons, whilst the epithelioid astrocytes upon which the neurons were growing are neither NSE-positive nor covered with silver grains. No NSE-positive neurons were seen that had accumulated I>aspartate after incubation times of 1-20 min (Fig. 4c, d).

In order to determine the stages during the combined autoradiography and intracellular staining procedure at which significant losses of accumulated [3H]GABA and D-[3H]aspartate occurred, cultures were solubilized after each step and the percentage of the radiolabel retained determined. Fig. 5 shows a

[3HI GABA ~-[3H]As~rto~o r e t e n t i o n 1~ r e t e n t i o n

[ ]0 min 3H-GABA or 3H-D-asPar ta te , lO0 lO0 m~ wash x4 t la+- f ree TBS

~ x 20 min 4 para fo rmaldehyde, 0 . ] ] 41.5 + 0.6 26.9 + l .O g l u t a r a l d e h y d e , wash x2 PBS I

i lO rain methanol ( - 2 0 o c ) , 32.7 + 0 . ] 19,4 + 0.9

I _ _ wash x4 ?BS - -

L 3O min ]0% normal goat serum in 25.5 + l . l 14.2 + 0.2 PBS, wash x4 PBS - -

L 30 min p r imary an t ibody in PBS ] 23.5 + ] . 4 12.7 + 0.5

+ 5 normal goat serum, wash x4 PBS J

3D min goat a n t i - r a b b i t IgG ] 21.3 ± 1.5 ]0.D + O in PBS, wash x4 PB$ ~ '~

,zac, h ~;) ,v/r:

Dehydrate i n graded e thano l se r i es 20.2 + 0.5 10,6 + 0 .2 and a i r dry - -

Fig. 5. Flow chart of the combined autoradiography and intracellular staining procedure and the corresponding retention of the accumulated [3H]GABA and o-[3H]aspartate after the various stages. The 20-DIV cultures, prepared from neonatal mouse forebrain, were incubated with the [3H]amino acids (10 -7 M), the glass coverslips placed in scintillation vials, and the tissue solubilized with 0.5 N NaOH after each of the steps above. Values are means + S.E.M. from 9 cultures from 3 independent experiments.

PROCEDURE

flow chart of the procedure and the corresponding percentage retention after each stage. As expected the greatest loss occurred during the fixation, 58.5% and 73.1% for [3H]GABA and D-[3H]aspartate respectively. The next greatest loss occurred during the methanol permeabilization step, 8.8% and 7.5% respectively. Radiolabel was gradually lost during all the subsequent steps resulting in 20.2% [3H]GABA and 10.6% D-[3H]aspartate remaining at the end of the procedure. It is clear that any cells that were only sparsely covered with silver grains after fixation with 2% glutaraldehyde would probably no longer be labelled after the whole procedure outlined in Fig. 5. However, in the present study the neurons and process-bearing glial cells that were labelled with [3H]GABA and D-[3H]aspartate after fixation with

2% glutaraldehyde were still labelled, although less intensely, after the whole intracellular staining pro-

cedure.

DISCUSSION

In the study reported here we have shown that individual [3H]GABA and D-[3H]aspartate accumulating cells could be reliably identified by the simultaneous use of immunocytochemistry with cell type- specific antigenic markers and tritium autoradiography. The results show that many cells that had accumulated the neuroactive amino acids GABA and D- aspartate could be definitively identified as oligodendrocytes by anti-GC immunofluorescence, thus con- firming our previous results that had been based on indirect methods 2°'21. A small number of the multi-

processed cells, growing on top of the GFAP-positive astroglial carpet layer, that were heavily labelled with GABA or D-aspartate, were also found to be GFAP-positive astrocytes, illustrating the usefulness of this method in identifying morphologically similar cells. Similar results were obtained in a previous study 1 using combined GFAP immunofluorescence and GABA autoradiography, although the method, using 2.5% glutaraldehyde as fixative, was not ex- tended to other antigens or transmitters.

A population of NSE-positive neurons in the neuron-enriched cultures expressed an avid uptake of GABA as expected TM. However, no D-aspartate-accumulating neurons could be identified, the silver grains occurring only over glial cells, in agreement

10

with other studies using dissociated cultures 5'26. In a

previous study 2 with retinal cell cultures it p roved

possible to identify neurons that were double label led

with [3H]GABA and tetanus toxin, using immunoflu-

orescence and benzoquinone as fixative. However

this method was found not to be satisfactory with D-

aspartate , and other antigenic cell markers were not

investigated. When using combined p e r o x i d a s e - a n -

t iperoxidase staining for g lu tamate decarboxylase

( G A D ) and G A B A autorad iography , Neale et al. TM

found it necessary to photograph and analyse the im-

munohis tochemistry pr ior to coating the slides with

emulsion and processing for au torad iography , in or-

der to rel iably identify double- label led neurons.

Otherwise only the cells that are not double- labe l led are clearly identif iable s,17.

The combined immunofluorescence and autora-

diography procedure described here has the advan-

tage that the f luorescence can be seen much clearer

through a covering of silver grains than the orange-

brown peroxidase react ion product (using diamino-

benzidine as substrate) . When using both surface and

intracel lular markers that give only a weak fluores-

cence, or only a dot ted broken pat tern , there is the

danger that the fluorescence would be obscured by

the overlying silver grains. In these cases it would be

necessary to adjust the silver grain density to the op-

t imum required to readily observe double- labe l led

cells, by varying the incubat ion t ime with the 3H-

t ransmit ter , the exposure t ime of the au torad iogram,

or the developing time. By changing the above vari-

ables and adapt ing the fixation and permeabi l iza t ion

procedures to the individual requirements of the anti-

gen to be localized, which may be achieved in part by

varying the g lutara ldehyde concentrat ion and the fix-

ation t ime, this method should be appl icable to a

wide variety of antigens 23 and radio label led neuro-

transmitters. Thus this technique should have a wide

applicabil i ty in determining whether selective neuro-

t ransmit ter t ranspor t mechanisms coexist with anti-

gens characterist ic of a certain cell type or subtype, in

individual cells, both neurons and glia. One recent

repor t 9 of the localization of [3H]muscimol and G A D

in retinal tissue sections suggests that this approach

may also be useful for use with tissue sections in addi-

tion to cultured cells.

ACKNOWLEDGEMENTS

This work was suppor ted by grants from the Swiss

Nat ional Science Founda t ion (3.493.83), the Swiss

Mult iple Sclerosis Society, and the Swiss Founda t ion

of Encouragement of Research in Menta l Re ta rda-

tion. The authors wish to thank Prof. M. Schachner

for helpful suggestions concerning the me thod and

Mrs. C. Steffen for technical assistance.

REFERENCES

1 Aloisi, F., Ciotti, M.T. and Levi, G., Characterization of GABAergic neurons in cerebellar primary cultures and selective effects of a serum fraction, J. Neurosci., 5 (1985) 2001-2008.

2 Beale, R. and Osborne, N.N., Selective uptake of tritiated glycine, GABA and D-aspartate by retinal cells in culture: a study using autoradiography and simultaneous immunofluorescence, Dev. Brain Res., 7 (1983) 107-120.

3 Burgess, S.K., Trimmer, P.A. and McCarthy, K.D., Auto- radiographic quantitation of fl-adrenergic receptors on neural cells in primary cultures. II. Comparison of receptors on various types of immunocytochemically identified cells, Brain Res., 335 (1985) 11-19.

4 Currie, D.N. and Dutton, G.R., [3H]GABA uptake as a marker for cell type in primary cultures of cerebellum and olfactory bulb, Brain Res., 199 (1980) 473-481.

5 Gordon, R.D. and Balazs, R., Characterization of sepa- rated cell types from the developing cerebellum: transport of glutamate and aspartate by preparations enriched in Pur- kinje cells, granule neurones, and astrocytes, J. Neuro- chem., 40 (1983) 1090-1099.

6 Haber, B. and Hutchison, H.T., Uptake of neurotransmit-

ters and precursors by clonal cell lines of neural cell origin, Adv. Exp. Med. Biol., 69 (1976) 179-198.

7 Hansson, E., Accumulation of putative amino acid neurotransmitters, monoamines and D-Ala-Met-enkephalina- mide in primary astroglial cultures from various brain areas, visualized by autoradiography, Brain Res., 289 (1983) 189-196.

8 Hatten, M.E., Francois, A.M., Napolitano, E. and Roff- ler-Tarlov, S., Embryonic cerebellar neurons accumulate 3H-GABA: visualization of developing GABA-utilizing neurons in vitro and in vivo, J. Neurosci., 4 (1984) 1343-1353.

9 Hendrickson, A., Ryan, M., Noble, B. and Wu, J.-Y., Co- localization of [3H]muscimol and antisera to GABA and glutamic acid decarboxylase within the same neurons in monkey retina, Brain Res., 348 (1985) 391-396.

10 H6sli, L. and H6sli, E., Action and uptake of neurotransmitters in CNS tissue culture, Rev. Physiol. Biochem. Phar- macol., 81 (1978) 135-188.

11 Iversen, L.L., Role of transmitter uptake systems in synap- tic neurotransmission, Br. J. Pharmacol., 41 (1971) 571-591.

12 Iversen, L.L. and Bloom, F.E., Studies of the uptake of3H - GABA and 3H-glycine in slices and homogenates of rat

brain and spinal cord by electron microscopic autoradiography, Brain Res., 41 (1972) 131-143.

13 Iversen, L.L. and Schon, F.E., The use of autoradiographic techniques for the identification and mapping of transmitter specific neurons in CNS. In: A. Mandell (Ed.), New con- cepts in Neurotransmitter Regulation, Plenum, New York, 1973, pp. 153-193.

14 Katz, D.M. and Kimelberg, H.K., Kinetics and autoradiography of high affinity uptake of serotonin by primary astrocyte cultures, J. Neurosci., 5 (1985) 1901-1908.

15 Kimelberg, H.K. and Katz, D.M., High affinity uptake of serotonin into immunocytochemically identified astrocytes, Science, 228 (1985) 889-891.

16 Lasher, R.S., The uptake of 3H-GABA and differentiation of stellate neurons in cultures of dissociated postnatal rat cerebellum, Brain Res., 69 (1974) 235-254.

17 Mazurkiewicz, J.E. and Messer, A., Simultaneous determi- nation of Leu-enkephalin localization and 3H-v-aminobutyric acid uptake in rat striatal cell cultures, Cell. Mol. Neu- robiol., 3 (1983) 255-262.

18 Neale, E.A., Oertel, W.H., Bowers, L.M. and Weise, V.K., Glutamate decarboxylase immunoreactivity and [3H]aminobutyric acid accumulation within the same neurons in dissociated cell cultures of cerebral cortex, J. Neuro- sci., 3 (1983) 376-382.

19 Reisert, I., Jirikowski, G., Pilgrim, C. and Tappaz, M.L., GABAergic neurons in dissociated cultures of rat hypothal- amus, septum, and midbrain, Cell Tissue Res., 229 (1983) 685-694.

20 Reynolds, R. and Herschkowitz, N., Uptake of [3H]GABA by oligodendrocytes in dissociated brain cell culture: a corn-

11

bined autoradiographic and immunocytochemical study, Brain Res., 322 (1984) 17-31.

21 Reynolds, R. and Herschkowitz, N., Selective uptake of neuroactive amino acids by both oligodendrocytes and astrocytes in primary dissociated culture: a possible role for oligodendrocytes in neurotransmitter metabolism, Brain Res., 371 (1986) 253-266.

22 Reynolds, R. and Herschkowitz, N., Involvement of oligodendrocytes in amino acid neurotransmitter metabolism. In T. Grisar, G. Franck, L. Hertz, W.T. Norton, M. Sensen- brenner and D. Woodbury (Eds.), Dynamic Properties o f Glial Cells. Cellular and Molecular Aspects, Pergamon, Ox- ford, 1986, pp. 255-264.

23 Reynolds, R. and Herschkowitz, N., Oligodendroglial and astroglial heterogeneity in mouse primary CNS culture as demonstrated by differences in GABA and D-aspartate transport and immunocytochemistry, Dev. Brain Res., 36 (1987) 13-25.

24 Schmechel, D.E., Brightman, M.W. and Barker, J.L., Localization of neuron-specific enolase in mouse spinal neurons grown in tissue culture, Brain Res., 181 (1980) 391-400.

25 Semenoff, D. and Kimelberg, H.K., Autoradiography of high affinity uptake of catecholamines by primary astrocyte cultures, Brain Res., 348 (1985) 125-136.

26 Wilkin, G.P., Levi, G., Johnstone, S.R. and Riddle, P.N., Cerebellar astroglial cells in primary culture: expression of different morphological appearences and different ability to take up [3H]D-aspartate and [3H]GABA, Dev. Brain Res., 10 (1983) 265-277.

simultaneous immunofluorescence and autoradiography: a useful technique for investigating...

Documents