GLIA 6:108-117 (1992)

Expression of Adrenergic Receptors in Individual Astrocytes and Motor Neurons Isolated From the Adult Rat Brain YANPING SHAO AND JEROME SUTIN Department of Anat0m.y and Cell Biology, Ern0r.y University School of Medicine, Atlanta, Georgia 30322

KEY WORDS

Glia, Synapse efficacy, Cortex, Cerebellum, Striatum, Trigeminal, Degeneration

ABSTRACT

Attempts to show the distribution of adrenergic receptors ( A R s )in autoradiographs of a brainstem motor nucleus following elimination of motor neurons yielded the unexpected result of a n increase in P-AR density. This increase was related to the gliosis accompanying the motor neuron degeneration. To determine the cells on which the AFi subtypes were located, we dissociated cells from various regions of the adult rat brain and subsequently identified astrocytes by glial fibrillary acidic protein (GFAP) immunofluorescence. Slides containing the astrocytes were prepared for autoradiography using the nonselective p ligand 1251-iodocyanopindolol(l2'IICYP) or the a1 ligand blocker betaxolol or the p2 '"IBE 2254 (l"I-HEAT). The addition of the selective blocker ICI 118.551 to the incubation medium to displace 1"51CYP binding was used to determine the binding of P-AR subtypes. The great majority (>88%) of isolated astrocytes sampled from the trigeminal motor nucleus, cerebral cortex, striatum, and cerebellum showed p-AR binding. Astrocytes from the first three regions had similar average densities of P-ARs, whereas the density in cerebellar astrocytes was 2- to 3-fold greater. The &-AR subtype was proportionally greater than the PI subtype m each region. Reactive astrocytes isolated from the trigeminal motor nucleus after degeneration of motor neurons showed a p-AR density nearly 2-fold greater than resting astrocytes from the same region, with the P1 subtype showing the greater proportional increase. There was no p-AR binding on trigeminal motor neurons. Astrocytes also showed a significant level of ol,-AR binding. No differences in a,-AR binding were found in normal astrocytes isolated from the different regions, nor was there a n increase in reactive astrocytes. In contrast, trigeminal motor neurons had a n a,-AFidensity nearly 10 times greater than astrocytes. In terms of the NE modulation of synaptic responses in motor neurons, the distribution of A R s would permit NE to act indirectly through a1 and p receptors on astrocytes and directly through a1 receptors on motor neurons. (c 1992 Wiley-Liss, Inc.

INTRODUCTION Norepinephrine (NE) enhances the response of spinal motor neurons (Fung and Barnes, 1987: White and Neuman, 19831, ceretellar Purkinje'cells (Woodward et and hippocampa' (Segal and 1976) following synaptic activation or iontoDhoretic application of excitatory or inhibitory amino acids. I n 1978)7

-"

0 1992 Wiley-Liss, Inc.

I

trigeminal motor nucleus (MoV), NE facilitates the motor neuron responses evoked from Ia afferent axons (the

xceivedFt.bruary28, 1991;accepted December

,991,

Y. Shao is now at the Department of Pharmacology, University ofNorth Carolina School of Medicine, Chapel Hill. NC 27599-7365. Address reprint requests to Dr. J. Sutin at the address given above.

GLIAL ADRENERGIC RECEPTORS IN ADULT BRAIN

109

masseteric reflex), which is partially blocked by either trypsinization was then terminated by incubation with al- (Stafford and Jacobs, 1990) or p- (Vornov and Sutin, 0.1% soy bean trypsin inhibitors (Sigma). Some tissue 1986) adrenergic receptor (AR) antagonists. Intracellu- from the cerebellum was incubated in the absence of lar recordings from motor neurons indicate that the trypsin to assess the effect of the enzyme. Following the increase in EPSPs may be due to a presynaptic change, trypsin treatment, solutions were kept on ice throughpresumably a n increase in the amount or duration of out the rest of the procedure. transmitter release (Vornov and Sutin, 1986). To determine the site of NE actions, it is important to determine Combined Immunofluorescence and if ARs are located solely on motor neurons or also on Receptor Autoradiography other cells that could indirectly mediate the facilitation. Autoradiographic studies in brain sections provided The combined method for GFAP immunoreactivity strong but indirect evidence for the expression of adrenergic receptors in both resting and reactive glial cells and adrenergic receptor autoradiography developed by (Shao and Sutin, 1990). Receptor binding autoradiogra- McCarthy (19831, modified by dipping rather than dryphy showed a decrease in a,-ARs associated with motor mount apposition of the emulsion (Katz and Kimelberg, neuron degeneration and a marked increase in p-ARs, 1985), was used to measure AR binding on individual which coincided with the glial reaction induced by the identified astrocytes. Cells smeared on glass slides were fixed for 10 min in injection of Ricin communis into the masseter nerve. 10% formalin to yield better cell morphology under Since astrocytes regulate extracellular potassium and take up neurotransmitters, including monoamines (Se- phase-contrast optics. Following 2 rinses in 0.05 M menoff and Kimelberg, 1985), the presence of A R s may PBS, cells were incubated for 30 min with a monoclonal imply a role of these cells in the presynaptic facilitation antibody against glial fibrillary acidic protein (GFAP) of motor neuron response to Ia afferent inputs by NE (Accurate Chem. Co.) diluted 1:lOO in PBS containing (Vornov and Sutin, 1986). However, autoradiography in 0.1% saponin, followed by exposure to a secondary goattissue sections does not resolve receptor binding a t the antimouse IgG antibody conjugated to rhodamine (Accurate Chem. Co.). Omission of the anti-GFAP antibody single cell level. Astrocytes in primary culture express ARs (Burgess resulted in absence of staining. Following GFAP immuet al., 1985; Lerea and McCarthy, 1989). p-ARs were nostaining, cells were washed briefly in 0.17 M Trisobserved in astrocytes isolated from cerebral cortex of HCl buffer (pH 7.6) containing 10 mM MgC1, and 0.01% the adult rat (Salm and McCarthy 1989), but the sub- ascorbic acid. For p-AR binding, the immunostained cells were intypes of p-ARs and their existence in astrocytes isolated from other brain regions is not known. To answer these cubated in the Tris-HC1 buffer containing 30 pM lZ5Ioquestions, single normal and reactive astrocytes and docyanopindolol (1251CYP) (2,200 Cdmmole, Amerneurons were isolated from various regions of the adult sham). 1251CYP has been characterized to bind specifically to p-ARs (Engel et al., 1981; Hoyer et al., rat brain and analyzed by receptor autoradiography. 1982). With a Kd of 6 pM for brain tissues (Sutin and Minneman 19851, 30 pM should theoretically bind 83% METHODS of receptors. To control for nonspecific binding, sepaBulk Dissociation of Neurons and Glial Cells rate slides were incubated in l pM dL-propranolol Cells were isolated from the brains of 2- to 6-month- (Sigma) or 200 pM -+-isoproterenol (Sigma) together old Sprague-Dawley rats by the bulk dissociation with 1251CYP.To assess p-AR subtypes, 200 nM betaxmethod of Farooq and Norton (1978). After removing olol, a p1 blocker, was added to restrict lZ51CYPbinding the meninges, tissue from the cerebral cortex, striatum, to the pz subtype; 200 nM ICI 118,551, a 6, blocker and cerebellum was dissected into 1mm3 pieces. Addi- (Bilski et al., 1980), was added to limit 1251CYPbinding tional tissue was taken by 1 mm diameter punches of to the PI subtype. (Both the and p, blockers were the trigeminal motor nucleus (MoV) in transverse sec- kindly provided by Dr. Kenneth P. Minneman.) Betaxtions of the pons. In order to study reactive astrocytes, olol has a high affinity (Kdl 7 nM) for the p,-AR and a gliosis was induced in MoV on one side of the brain by low affinity (h2 - 2,000 nM) for the p,-AR in competiselectively destroying motor neurons. Under ketamine tion with 1251CYP(40-50 pM) (Engel e t al., 1981). Conanesthesia (50 m g k g body weight, i.m. diluted 1:l in versely, ICI 118,551 displays a high affinity -7 xylazine), the masseter nerve branch of the left trigem- nM) for the p,-AR and a low affinity (Kd2 3,500 nM) inal motor root was exposed by removing the zygomatic for the p,-AR. For a,-ARbinding, isolated cells were incubated with arch. One pl of saline containing 1.5 pg of the retrogradely transported cytotoxic lectin Ricin communis 80 pM of the a, antagonist lZ5IBE 2254 ('"I-HEAT), (Sigma) was injected into the nerve. Control rats re- which was a Kd value ranging from 32 to 61 pM (Engel ceived a n injection of saline only. Most of the animals et al., 1981; Lerea and McCarthy, 1989; Minneman, were allowed to survive for 1 4 weeks after the injec- 1983). Nonspecific binding was determined by incubation in the presence of 1 pM prazosin (Sigma) or 50 pM tion. Tissue samples were incubated in 37°C Ficoll solu- phentolamine (Sigma) in the incubation solution contion containing 0.1% trypsin (Sigma) for 60 min. The taining lZ51-HEAT.

-

-

110

SHAO AND SUTIN

The incubations lasted for 2 h and were followed by TABLE I . Yield of GFAPpositiue cells isolated from adult rat braina two 20-min washes in cold Tris-HC1 buffer and 5 min in Brain Sample Total GFAP+ GFAP + as weight (ma) cells/mm2 cells/mm2 distilled deionized water to remove salts. Cells were region %I total cells then dried in cold air and stored overnight with desic- Neocortex 800 280 k 9.6 5.0 t 1.3 1.8 500 123 k 25 1.2 t 0.8 1.0 cant at 4°C. The slides were coated with Kodak NTB-3 Striaturn 46.6 k 6.1 3.1 1,000 1514 k 70 Cerebellum emulsion by dipping in a n 1 : l dilution maintained a t MoV 57 i 5.3 0.6 k 0.4 1.1 150 45°C. Negligible tissue chemography and edge effect Ricin MoV 80 76 k 8.2 6.8 k 1.6 8.9 were observed in slides processed without the ligand "Cell density was determined after immunostaining and autoradiography. Cells incubation step. Emulsion-coated slides were dried and were counted in a 0.5 mm2field at a magnification of 400X under phase-contrastto determine total cells, and then under epifluorescence to identify GFAP+ cells. For stored with desiccant in lightproof boxes a t - 70°C for 2 each brain region, 20 separate fields in 2 slides were counted and expressed a s mean weeks. The emulsion was developed in Dektol(1:l) for 2 & s.e.m.. Among the normal tissue samples, cerebellum yielded more GFAPf cells than others. Compared to normal MoV, the ricin-treated MoV (Ricin MoV) showed min, fixed in Kodak fixer for 5 min, and washed in an 8-fold increase in GFAP+ cells. distilled water for 20 min. Slides were then dried, coverslipped with Entellan (EM Co.), and stored at 4°C. Cells were examined under phase-contrast and epi- No further purification of the isolated cells was underfluorescence with a n Olympus VANOX microscope, and taken for the following reasons: (1)the shear force of astrocytes identified by GFAP immunoreactivity. Only high speed centrifugation might damage cells and resingle cells not contacting others were analyzed. Since duce receptor binding on cell surfaces (Guarnieri et al., it is difficult to accurately measure the cell surface area 1975); (2) the quantity of tissue required to obtain puriof astrocytes due to their irregular shape, binding den- fied astrocytes was too large (>1g) to obtain from MoV sity was estimated by counting the silver grains in a tissue ( 0.05), which differed from that and McCarthy (1989). The total '"'ICYP binding in the of cerebellar astrocytes (F,,,,,,, = 48.5, P < 0.001). different regions was 2.6- to 6.5-fold higher than the nonspecific binding ( P < 0.001). Whereas astrocytes isolated from the cerebral cortex, striatum, and MoV showed similar average grain densities, the density P-AR Subtypes over cerebellar astrocytes was 2- to 3-fold greater. The To determine the subtype of p-ARs on astrocytes, the effect of the trypsin treatment used in the dissociation procedure was tested in 1251CYPbinding assays of cells selective p1 blocker betaxolol or the p2 blocker ICI isolated from the cerebellar cortex. The mean and stan- 118,551 was used to displace the specific binding of dard error of specific binding in GFAP positive cells not 1251CYP.At the concentration used, the blocking drugs exposed to trypsin was 52 k 4.3 silver grains per 1,000 should be selective for each subtype, and a proportional

*

112

SHAO AND SUTIN

Fig. 2. '"ICYP binding in astrocytes isolated from the adult rat brain. Two protoplasmic astrocytes from the cerebellum shown with phase-contrast (A) and with combined epifluorescence and darkfield optics (B) are labelled with silver grain clusters. High densities of silver grains are also associated with astrocytes from the cerebral cortex (C), striatum (D), MoV (E), and ncin-treated MoV (F). In contrast, trigeminal motor neurons (G) show only a background level (H).Bar = 50 km.

displacement (9% p1 + % p2 = 100% p total) would be expected. As shown in Table 2, the sums of the p1 and P2 binding ranged between 83% and 105% of the total p binding. The percentage for each region was calculated by taking the sum of each subtype binding, subtracting nonspecific binding and dividing by the total I2'1CYP binding minus nonspecific binding. Cortical, striatal, and MoV astrocytes had similar levels of Pa (P < 0.001) and p1 binding (0.05 > P >

0.01). Cerebellar astrocytes had higher levels of both p1 and pz binding ( P < 0.001); 73-98% of astrocytes in all samples exhibited pz binding (Figs. 3 , 4 ) .

P - A F t s on Reactive Astrocytes

Reactive astrocytes isolated from MoV after neuronal degeneration were associated with silver grain clusters

GLIAL ADRENERGIC RECEPTORS IN ADULT BRAIN TABLE 2. Specific binding of 0-AR subtypes on astrocytes from different brain regionsa

113

p-ARs. We confirmed their observations and also found

p-ARs in astrocytes isolated from the striatum, trigeminal motor nucleus (MoV), and cerebellar cortex. The cell isolation procedure, especially the enzyme 16 f 1.6 13 1.9 3 + 1.7 Neocortex (651 (55) (49) treatment, has the potential to destroy membrane re15'+ 2.0 6 k 1.6 25 & 1.7 Striatum ceptors. Guanieri et al. (1975) reported that opiate re(75) (60) (50) ceptors, but not cholinergic muscarinic receptors, were 12 + 1.3 46 i 3.3 55 f 3.0 Cerebellum (114) (50) (50) destroyed by trypsin. Control experiments in which MoV 25 + 3.4 15 + 1.9 5 zk 1.5 binding in nontrypsin-treated, isolated cerebellar as(55) (45) (40) trocytes was compared with that in cells exposed to the Ricin MoV 45 z-2.7 26'+ 2.5 18'+ 2.1 (75) (55) (55) enzyme, as well a s data reported by Salm and McCarthy (1989), indicate that p-AR binding is not altered by " D adrenergic receptor binding of '*sIodocyanopindolol (ICYP) alone, p-2 binding ('"ICYP in the presence of betaxolol), and p-1binding ('251CYPin the presence of the concentration of trypsin used in our study. ICI~l18,551).All values expressed a s silver grain density (mean f s.e.m./1,000 pmz) minus nonspecific binding. Numbers in parentheses = number of astrocytes Although isolated cells free of obvious cellular debris sampled. were selected for analysis, it is not absolutely certain that membrane components of other cell types were not associated with astrocytes or motor neurons. In a similar preparation, Salm and McCarthy (1989) double-ladue to 12'ICYP binding. Compared to resting astrocytes belled cells with antibodies to GFAF' and neurofilament isolated from normal MoV, reactive astrocytes had or galactocerebroside to demonstrate that isolated asnearly twice the density of total p binding (Table 2), trocytes were not contaminated with adherent compowith the subtype increasing proportionally more nents from neurons o r oligodendrocytes. Higher densithan p,-ARs (Fig. 5). binding density was 3.7 times ties of silver grains occurred over astrocytes but not greater and less variable in reactive astrocytes. The over tissue debris, indicating that adherent membrane difference in average silver grain density in resting and fragments from disrupted cells do not contribute signifreactive astrocytes was significant at the P < 0.01 level icantly to labelling of ARs on astrocytes. (unpaired t-test). Silver grain density was measured in a standard-size window positioned to cover the soma and proximal dendrites, but that included cell-free areas between procw,-ARs o n Trigeminal Motor Neurons cesses, and the results normalized to density per unit area (1,000 pm2). Since this differs from density per In contrast to astrocytes, motor neurons isolated unit area of membrane surface, it tends to underestifrom MoV showed marked labelling with '"1-HEAT mate the true density of silver grains associated with (Fig. 6A,B). The nonspecific binding in the presence of cells. The validity of these measurements was estab50 pM phentolamine or 1 pM prazosin was generally lished by manually counting silver grains confined to high in all cells, especially in motor neurons (Table 3). the soma and proximal processes of several astrocytes. It was not determined if part of the nonspecific binding Another source of underestimation was due to the fuwas due to incomplete blockage of the cw,-ARs. Never- sion and superimposition of multiple silver grains. The theless, the specific a1binding in trigeminal motor neu- division of total grain area by average grain size parrons was nearly 10 times greater than that in astro- tially corrected for fusion, but not for superimposition. cytes from the same region. Astrocytes have been historically classified into a t least two types, described as fibrous and protoplasmic. Raff et al. (1984) classified astrocytes in the rat optic cw,-ARs on Astrocytes nerve and primary tissue cultures into type 1 and type 2, based on their distinct morphology and cell surface To determine a,-AR binding, isolated cells were incu- antigens. I n primary cultures, type 1 astrocytes are bated in "'I-HEAT. An example of a n astrocyte is polygonal and have a high density of P A R S , whereas shown in Figure 6C,D. Astrocytes isolated from all re- type 2 astrocytes are process bearing with few p-AR gions examined showed low but significant labelling, binding sites (Burgess and McCarthy, 1985; Burgess et with the mean densities of specific binding ranging al., 1985). The relation between rat optic nerve type 1 from 13 to 19 grains/1,000 pm2 (Table 3). One-way and type 2 in culture, and protoplasmic and fibrous analysis of variance showed no differences among as- astrocytes in most brain regions in vivo has not been trocytes from the various brain regions, including established (Miller e t al., 1989). Isolated cerebellar asreactive astrocytes from the ricin-treated MoV trocytes have a P - A R density substantially greater than (F,,,,,,, = 1.2, P > 0.05). astrocytes from the cerebral cortex, striatum, and MoV. The difference in astroglial expression of P-ARs is not related to NE levels in the brain, since the NE content follows the order of MoV > cerebral cortex > cerebellum DISCUSSION > striatum (Jonsson and Sachs, 1982; Sutin and MinSalm and McCarthy (1989) demonstrated that astro- neman, 1985). In keeping with greater p - A R density in cytes isolated from adult rat cerebral cortex express cerebellar astrocytes, NE stimulated CAMP levels are Brain region

Total 0-AR binding

el

P2-AR binding

*

01-AR binding

SHAO AND SUTIN

114

Distribution of Beta- AR Binding on Astrocytes from Different Brain Regions

30

__ Cortex

fn Q)

c

h

20

0

-.-

Striatum

-----

Cerebellum

e

c

a

MeV

a aR

10

0

o

20

40 60 ao 100 1 2 0 Silver Grains11 ,000 urn

140

Fig. 3. Relative frequency of astrocytes plotted as a function of silver grain densities (bin width = 10). The density of p-AR binding was determined by the subtraction o f the nonspecific binding from the total binding (Table 2); 12% o f cortical, 8% of striatal, 6%of MoV, and 2% of cerebellar astrocytes sampled do not show p-AR binding. The distributions are similar in the former 3 regions, whereas cerebellar astrocytes frequently have a higher density and a greater variation in p - A R expression.

Distribution of Beta-2 Binding on Astrocytes from Different Brain Regions 6o

r

50 ~

40

cn Q)

x 0

Y

e 2

30

Cortex

-.-

Striatum

-----

Cerebellum

c

20

MeV

10

0

Fig. 4. Distributions of astrocytes expressing the pz subtype. The graphs were plotted as described in Figure 3. 22% of the cortical, 27% of striatal, 18%o f MoV, and 3% of cerebellar astrocytes show no p2 binding. Similar to the total 8-AR binding, cerebellar astrocytes show a higher density and a greater variation in p,-AR expression.

greater in glial enriched fractions from this region (Palmer, 1973). The fact that p-ARs are not found on motor neurons supports the argument, based on intracellular record-

ings, that the p-AR mediated facilitation of the trigeminal motor neuron monosynaptic reflex is due to a presynaptic site of action (Vornov and Sutin, 1986). The afferent axons of this reflex path release an excitatory

115

GLIAL ADRENERGIC RECEPTORS IN ADULT BRAIN

Proportion of Beta- AR Subtypes on Isolated Astrocytes

CD

C

p.-

a

60.

(D c Q)

a

0

I Cortex

Striaturn

Beta-1

Cereb.

MoV

Ricin MoV

Beta- 2

Fig. 5. The proportion of p-AR subtypes on isolated astrocytes. The binding of ‘“‘ICYP was displaced by the blocker betaxolol or the pz blocker ICI 118,551 separately. The p subtype binding was determined by the subtraction of nonspecific binding from the total subtype binding (Table 2). The average density of each subtype binding was

compared to that of the total p binding to yield a percentage. This graph shows that the sums ofthe two subtypes range from 83-107% of the total (% + % pz = 100%p total), suggesting a proportional displacement. In all regions examined, the portions of the Pa binding are greater than those of the p1binding.

amino acid, probably glutamate, which acts through non-NMDA receptors on the motor neuron (Chandler, 1989). The known properties of astrocytes include several features that could play a role in the presynaptic modulation of motor neuron excitability. Astrocytes are closely apposed to motor neurons (Peters et al., 1979) and express P-ARs. Activation of P-ARs on cultured astrocytes causes a hyperpolarization that is not associated with a change in membrane resistance (Hosli et al., 1982)and influences the active uptake of the amino acid neurotransmitters glutamate and GABA (Hansson and Ronnback, 1989). Evidence supporting the hypothesis that astrocytes influence neuronal activity has been reviewed by Vernadakis (1988).

Polygonal astrocytes in primary cultures express only the p1 subtype (Beaker et al., 1986). Nevertheless, the p2 subtype is found in human fetal astrocytes and mediates a substantial elevation of intracellular CAMP levels (Woods et al., 1989). It is not surprising that different astrocytes express different subtypes of receptors, since many factors can regulate expression of P-AR subtypes. Voisin et al. (1987) demonstrated that cultured cerebellar astrocytes express both and p2 subtypes. The p1 subtype is predominant when the astrocytes are grown in fetal calf serum in which they exhibit a polygonal, poorly differentiated morphology; when culture conditions favor cellular differentiation, the p2 subtype becomes predominant. Since pI-ARs are down-regulated in the presence of increased NE but p2-ARs generally are not, it is conceivable that in the brain pl-ARs are tonically suppressed by NE, so p,-ARs are more numerous, whereas in tissue culture pl-ARs are up-regulated.

p-AR Subtypes on Astrocytes The existence of both p1 and p2-ARs on astrocytes in the brain is supported by the displacement of 1251CYP binding with either the p1 blocker betaxolol or the pz blocker ICI 118,551. The proportional displacement by the two ligands suggests that each blocked a distinct subpopulation of p-ARs. Although the ligand 1251CYP has generally been reported to bind nonselectively t o both p1 and p2 subtypes (Engel et al., 19811, Neve et al. (1987) showed that at a concentration near &,1251CYP can have a 2-fold selectivity in favor of the p2subtype in C, glioma cells. If this is true for astrocytes in the brain, PIbinding would be underestimated more than p2binding.

Increase in P-ARs on Astrocytes During the Glial Reaction Compared with resting astrocytes, reactive astrocytes isolated from MoV exhibit a 2-fold higher density of p-ARs. It is uncertain if all the reactive astrocytes are derived from the normal (resting) GFAP positive astrocytes, since a population of GFAP negative precursor cells from the adult rat brain has been reported to become GFAP positive during proliferation (Norton et al.,

116

SHAO AND SUTIN

Fig. 6. Examples of '"I-HEAT binding on a motor neuron (A and B) and an astrocyte (C and D) isolated from MoV. A. Phase-contrast image of a trigeminal motor neuron with a triangular shape and two truncated proximal dendrites. B. ,4 darkfield image of A shows silver grains accumulated over the motor neuron. C . GFAF' immunofluorescence of an isolated astrocyte. D. Darkfield photomicrograph of the same cell illustrating the silver grains associated with the astrocyte. Bars = 50 IJ-m.

TABLE 3. al-AR binding on trigeminal motor neurons and astrocytes isolated f r o m adult rat brain"

Cell Type

Total '""I-HEAT binding

Trigeminal Motoneurons Trigeminal Astrocytes Cerebellar Astrocytes Cor t ica 1 Astrocytes Reactive Astrocytes

222 i 7.4 (48)' 40 iz 2.9 (36) 50 2.7 (36) 45 k 3.3 (30) 45 i 3.5 (30)

'251-HEAT+ '"I-HEAT+ Prazosin phentolamine (nonspecific)

*

107 ?c 7.2 (26) 28 c! 3.8 (18) 32 i 4.3 (18) 32 iz 2.9 (18) 26 3.4 (18)

*

89 3z 6.4 (30) 25 5 3.5 (20)

al-AR bindingb

* 7.4** 12 * 2.9*

115

18 f 2.7* 26 3z 4.3 (18)

13 i 3.3* 19 i 3.5*

,'All values expressed as the number of silver grains (mean f s.e.m.)/1,000pm2. '>Specificn,-AR binding = total '2sI-HILATbinding-nonspecific binding ('2511-HEAT + prazosin).

'Number of cells sampled. 'KP< 0.05. **P< 0.001. Student's unpaired t-test,comparing total "511-HEATbinding with nonspecific binding (l"I-HEAT

1988; Norton and Farooq, 1989). The increase in P - A R density in reactive astrocytes agrees with the findings in brain sections that the glial reaction in the MoV with motor neuron degeneration is associated with greatly increased p-AR density. Reactive astrocytes express high levels of pl-ARs, whereas resting astrocytes exhibit only a low level,

+ prazosin).

which may reflect a similarity between hypertrophy of these cells in vivo and the transformation from a flat to a process-bearing morphology in primary cultures (Shain et al., 1987). The NE level in the MoV with extensive gliosis remains unchanged (Shao and Sutin, 1990), so up-regulation of glial receptors is unlikely.

GLIAL ADRENERGIC RECEPTORS IN ADULT BRAIN

cy,-ARs on Astrocytes and Motor Neurons The radioligand lZ51-HEATlabels cu,-ARs in crude cerebral cortical membrane preparations (Engel and Hoyer, 1981; Minneman et al., 1983), brain sections (Jones et al., 1985), and heterogenously in cultured brain cells (Lerea and McCarthy, 1989). Our results demonstrate binding of lZ51-HEATin both isolated astrocytes and motor neurons, with the density very much greater in the neurons. The high density of alARs in motor neurons would explain the decreased lZ5HEATbinding observed in tissue sections through regions of motor neuron degeneration (Shao and Sutin, 1991).

ACKNOWLEDGMENTS This work was supported by National Institutes of Health grant NS 14778. We are indebted to Ronald Griffith for excellent technical assistance and to Dr. Kenneth Minneman €or kindly providing radiolabeled ligands.

REFERENCES Beaker, S., Sumner, C., Pitha, J., and Raizada, M.K. (1986) Characteristics of the p-adrenoceptor from neuronal and glial cells in primary cultures of rat brain. J . Neurochem., 47:131€&1325. Bilski, A., Dorries, S., Fitzgerald, J.O., Jessup, R., Tucker, H., and Wale, J. (1980) ICI 118, 551, a potent 0,-adrenoceptor antagonist. Br. J . Pharmacol., 69:292P. Burgess, S. and McCarthy, K.D. (1985)Autoradiographic quantitation of beta-adrenergic receptors on neural cells in primary cultures. I. Pharmacological studies of 1251-pindolol binding of individual astroglial cells. Brain Res., 335:l-9. Burgess, S.K., Trimmer, P.A. and McCarthy, K.D. (1985) Autoradio$aphic quantitation of beta-adrenergic receptors on neural cells in primary cultures. 11. Comparison of receptors on various types of immunocytochemically identified cells. Brain Res., 335:ll-19. Chalidler, S.C. (1989)Evidence for excitatory amino acid transmission between mesencephalic nucleus of V afferents and jaw-closer motorneurons in the guinea pig. Brain Res., 477:252-264. Engel, G. and Hoyer, D. (1981) '"'IBE 2254, a new high affinity radioligand for u,-adrenoceptors. Eur. J . Pharmacol., 73221-224. Engel, G., Hoyer, D., Berthold, R., and Wagner, H. (1981) Iodocyanopindolol, a new ligand for 6-adrenoceptors: Identification and quantitation of subclasses of 0-adrenoceptors in the guinea pig. Naunyn Schmiedebergs Arch. Pharmac., 387:277-285. Farooq, M. and Norton, W.T. (1978)A modified procedure for isolation of astrocytes- and neuron-enriched fractions from rat brain. J . Neurochem., 31:887-894. Fung, S.J. and Barnes. C.D. (1987) Coerulospinal enhancement of repetitive firing with correlative changes in postspike afterhyperpolarization of cat spinal motoneurons. Arch. Ital. Biol., 125:171-186. GuaTnieri, M., Krell, L.S., McKhann, G.M., Pasternak, G.W., and Yamamura, H.I. (1975) The defects of cell isolation techniques on neuronal membrane receptors. Brain Res., 93:337-342. Hansson, E. and Ronnback, L. (1989) Regulation of glutamate and GABA transport by adrenoceptors in primary astroglial cell cultures. Life S C ~ .4437-34. , Hosli. L.. Hosli. E.. Zehntner. C.. Lehmann. R.. and Lutz, T.W. (1982) Evidence for 'the' existence of alpha- and beta-adrenoceptors on cultured glial cells: An electrophysiological study. Neuroscience, 7:2867-2872. Hoyer, D., Engel, G., and Berthold, R. (1982) Binding characteristics of ( $: ), ( i) and ( - ) '2"Iodocyanopindolol to guinea-pig left ventricle membranes. Naunyn Schmiedebergs Arch. Pharmacol., 318:319. Jones, L.S., Gauger, L.L., and Davis, J.N. (1985) Anatomy of brain alpha,-adrenergic receptors: In vitro autoradiography with ""IHEAT. J . Comp. Neurol., 231:190-208.

117

Jonsson, G . and Sachs, C. (1982) Changes in the development of central noradrenaline neurons after neonatal axon lesions. Brain Res. Bull., 91641-650. Katz. D.M. and Kimelbere. H.K. (1985)Kinetics and autoradiography of high affinity uptake i f serotonin by primary astrocyte cultures. J . Neurosci., 5:1901-1908. Lerea, L.S. and McCarthy, K.D. (1989) Astroglial cells in vitro are heterogeneous with respect to expression of the ul-adrenergic receptor. Glia, 2:135-147. McCarthy, K.D. (1983) An autoradiographic analysis of beta adrenergic receptors on immunocytochemically defined astroglia. J . Pharmacol. Exp. Ther., 226:282-290. Miller, R.H., ffrench-Constant, C., and Raff, M.C. (1989b) The macroglial cells of the rat optic nerve. Annu. Reu. Neurosci., 12517-534. Minneman, K.P. (1983) Binding properties of a1-adrenergic receptors in the cerebral cortex: Similarity to smooth muscle. J . Pharmacol. Exp. Ther., 227:604-612. Minneman, K.P., Wolfe, B.B., Pittman, R.N., and Molinoff, P.B. (1983) p-adrenergic receptor subtypes in rat brain. In: Molecular Pharmacology of Neurotransmitters and Receptors. T. Segawa et al., eds. Raven Press, New York, pp. 73-81. NBve, K.A., McGonigle, P., and Molinoff, P.B. (1986) Quantitative analysis of the selectivity of radioligands for subtypes of p-adrenerpic receptors. J. Pharmacol. Exp. Ther., 238:46-53. Norton, W.T. and Farooq, M. (1989) Astrocytes cultured from mature brain derive from glial precursor cells. 1.Neurosci., 9:769-775. Norton, W.T., Farooq, M., Chiu, F-C., and Bottenstein, J.E. (1988) Pure astrocyte cultures derived from cells isolated from mature brain. Glia, 1:403-414. Palay, S. and Chan-Palay, V. (1974) Cerebellar Cortex Cytology and Organization. Springer-Verlag, Berlin, pp. 288-321. Palmer, G.C. (1973) Adenyl cyclase in neuronal and glial-enriched fractions from rat and rabbit brain. Res. Commun. Chem. Puthol. Pharmacol., 5:603415. Peters, A., Palay, S.L., and De F. Webster, H. (1976) The Fine Structure of the Nervous System: The Neurons and the Supporting Cells. Saunders, Philadelphia. Raff, M.C., Abney, E.R., and Miller, R.H. (1984)Two glial cell lineages diverge prenatally in the rat optic nerve. Dew. Biol.: 10653-60. Salm, A.K. and McCarthy, K.D. (1989) Expression of beta-adrenergic receptors by astrocytes isolated from adult rat cortex. Glia, 2:346352. Segal, M. and Bloom, F. (1976) The action of norepinephrine in the rat hippocampus. IV. The effects of locus coeruleus stimulation on evoked hippocampal unit activity. Brain Res., 107:513-525. Semenoff, D. and Kimelberg, H.K. (1985) Autoradiography of high affinity uptake of catecholamines by primary astrocyte cultures. Brain Res., 348:125-136. Shain, W., Forman, D.S., Madelian, V., and Turner, J.N. (1987) Morphology of astroglial cells is controlled by beta-adrenergic receptors. J . Cell Biol., 105:2307-2314. Shao, Y. and Sutin, J . (1991) Noradrenergic facilitation of motor neurons: Localization of adrenergic receptors in neurons and non-neuronal cells in the trigeminal motor nucleus. Exp. NeuroZ., 114:216227. Stafford, I.L. and Jacobs, B.L. (1990) Noradrenergic modulation of the masseteric reflex in behaving cats. I. Pharmacological studies. J . I Neurosci., 10:91-98. Sutin, J. and Minneman, K.P. (1985) u l - and p-adrenergic receptors are co-regulated during both noradrenergic denervation and hyperinnervation. Neuroscience, 14:973-980. Vernadakis, A. (1988) Neuron-glia interrelations. Int. Rev. Neurohiol., 30:149-224. Voisin, P.J., Girault, J.M., Labouesse, J., and Viratelle, O.M. (1987) p-adrenergic receptors of cerebellar astrocytes in culture: intact cells versus membrane preparation. Brain Res., 404:64-79. Vwrnov, J.J. and Sutin, J. (1986) Noradrenergic hyperinnervation of the motor trigeminal nucleus: Alterations in membrane properties and responses to synaptic input. J . Neurosci., 6:30-37. White, S.R. and Neurman, R.S. (1983) Pharmacological antagonism of facilitatory but not inhibitory effects of serotonin and norepinephrine on excitability of spinal motoneurons. Neuropharmacology, 22:489494. Waods, M.D., Freshney, R.I., Ball, S.G., and Vaughan, P.R.T. (1989) Regulation of cyclic AMP formation in cultures of human foetal astrocytes by P,-adrenergic and adenosine receptors. 1.Neurochem., 535364-869. Woodward, D.J., Moises, H.C., Hoffer, B.J., and Freedman, R. (1978) Norepinephrine modulation of Purkinje cell responses evoked by afferent stimulation and by microiontophoresis of amino acid neurotransmitters. In: Iontophoresis and Transmitter Mechanisms i n the Mamm.aZian Central Nervous System. R.W. Ryall and J.S. Kelly, eds. Elsevier, Amsterdam. ~~~