Brahl Research, 587 (19921 1-12 © 1992 Elsevier Science Publishers B.V. All rights reserved 110116-8993/92/$(15.110

1

BRES 18069

Research Reports

VIP receptor subtypes in mouse cerebral cortex: evidence for a differential localization in astrocytes, microvessels and synaptosomal membranes J e a n - L u c M a r t i n , D o u g l a s L. F e i n s t e i n *, N a i c h e n Yu, O l i v i e r Sorg, C o l e t t e R o s s i e r a n d P i e r r e J. M a g i s t r e t t i Instimt tit, Physiok~gie, Facultt ~de Mt~tlt,cbit; Unil'ersitt~de Lausanne, Lausanne (Switzerland)

(Accepted 14 April 1992)

Key words: Gila: Blood vessel: Synaptosome; Peptide; Cyclic AMP: Glycogen: Cerebral cortex

The binding characteristics of a monoiodinated form of vasoactive intestinal peptide (M-[t251]VIP) to the membranes of astrocytes, intraparenchymal microvessels and synaptosomes were analyzed in mouse cerebral cortex. Binding to astrocytes, studied in primary cultures, indicates the presence of a single class of high affinity binding sites with a K d of 3.3 nM and a Bm,,~ of 565 fmol/mg protein. The structurally related peptide secretin does not compete for sites labeled by M-[125I]VIP, In cultured astrocytes, VIP has been previously shown to promote glycogenolysis4t. Secretin, despite its lack of interaction with sites labeled by M-[ t2"sl]VIP, stimulates glycogenolysis with an ECsct of (}.5 nM. thus demonstrating the presence in astrocytes of functional secretin receptors independent from those for VIP. Trypsinization of the primary astrocyte cultures followed by replating as secondary cultures, reveals a second class of low affinity binding sites, with a K d of 41.3 nM and a Bin.~ of 881 fmol/mg protein, Secretin does not compete for this class of low affinity binding sites either. Binding of M-[12'~i]VIP to intraparenchymal microvessels reveals the presence of two classes of binding sites with K d of 1.4 and 30,3 nM, and Bm,~ of 7,1 and 73,8 pmol/mg protein, respectively. Similar to what is observed in primary or secondary astrocyte cultures, secretin does not interact with these sites. In this cell type VIP stimulates cAMP formation with an ECs0 of 18 nM, while secretin is ineffective, Finally, in agreement with previous reports in rat and guinea pig cerebral cortex, two classes of binding sites are observed in synaptosomal membranes: it high affinity class with a K,j of 4,9 nM and a B,.,~ of 316 fmol/mg protein, and a low affinity class with a Kd of 42,8 nM and :t B,.l~~ of 1578 fmol/mg protein, ht contrast to wh.'tt is observed in non-neuronal membranes, in synaptosomal membranes, secretin effectively competes for sites labeled by M-[ ~Z'~l]VIPwith an E f t , of approximately 150 nM, These results indicate that secretin may represent a useful toni to discriminate between nearontll and non-ncurtmal VIP blndintl sites, since it competes with M-[:'~'~I]VIPexclusively for the neuronal class of binding sites,

INTRODUCTION Vasoactive intestinal peptide (VIP) is a 28 amino acid peptide first isolated by Said and Mutt from porcine duodenum 3~. Like several other peptides originally isolated from the gastrointestina! tract, the presence of VIP was subsequently demonstrated within the central nervous system, where it is present in particularly high concentrations in the cerebral cortex tg. In this area of the brain, VIP is contained in a homogeneous population of radially oriented interneurons with minimal arborization in the horizontal plane 3t'~. These morphological characteristics confer to VIP intracortical neurons the capacity to exert local input-output

functions within cortical columns z~. Other observations that support a neurotransmitter role for VIP in the cerebral cortex are (i) its ability to be released by appropriate stimuli 27"28and (it) the fact that VIP exerts a number of effects on cell function. Thus VIP affects the excitability properties of identified neurons '~, stimulates cAMP formation '5''~'~ and promotes glycogenolySiS24. in view of this latter action, VIP has been proposed to play a role in the local regulation of energy metabolism 2t. Binding studies, aimed at the characterization of the properties of VIP receptors in the cerebral cortex, have been carried out by several groups on resuspended membranes and synaptosomes. The former preparation

Correspondence: PJ. Magistretti, lnstitut de Physiologie. 7. rue du Bugnon, 1005 Lausanne, Switzerland. Fax: 141) 21-313-28-65. * Present address: Department of Neurobiology, Cornell University Medical College, New York. NY 10021, USA.

contains neuronal as well as non-neuronal membranes originating from glial cells and cells of the microvasculature, while the latter preparation is highly enriched with neuronal membranes '~. In most studies, which were conducted in rat and guinea pig cerebral cortex, two binding sites, with high and low affinities have been reported 36"3~'4'-44. Furthermore both peptide histidine-isoleucine (PHi) and secretin, two peptides structurally related to VIP, competed for sites labeled by M-[ t ! ~ l ] V l P ~'37"4z'4~. O n the basis o f f u n c t i o n a l s t u d i e s , e v i d e n c e has h o w e v e r a c c u m u l a t e d o v e r the years, i n d i c a t i n g that V I P c a n exert c e l l u l a r a c t i o n s o n n o n - n e u r o n a l cells, such as a s t r o c y t e s a n d cells o f t h e i n t r a p a r e n c h y m a l m i c r o v a s c u l a t u r e . T h u s , V I P has b e e n s h o w n to p r o m o t e glycogenolysis in p r i m a r y c u l t u r e s o f rat a n d m o u s e c e r e b r a l cortical astrocytes 2''4' a n d to s t i m u l a t e

cAMP formation in the same cell type 4'4x and in intraparcnchymal microvessels ~¢''~7. Furthermore, the binding characteristics of M-[~:'sl]VIP to large cerebral arteries have been reported 4t', and VIP binding sites have been localized by autoradiography in astrocyte cultures of rat CNS ~. In vitro autoradiographic studies have indicated a laminar distribution of VIP binding sites in the cerebral cortex, with the highest density observed it, layers I, II, IV and VI I''", However, since film autoradiography does not provide localization with a resolution at the cellular level, VIP binding sites on non-neuronal cells could not he studied in detail, In considering the foregoing set of heterogeneous observations, it seemed useful to characterize pharmacologically and kinctically the properties of VIP receptors in astrocytes and intraparenchymal microvessels ~md to compare them to those of synaptosomal membranes within the same species and the same area of the brain, For this reason we have conducted such a study in the mouse cerebral cortex using primary astrocyte cultures, intraparenchymal microvesscis and synaptosomes. Results reported in this article indicate the presence of three distinct VIP receptor subtypes differentially localized on neuronal and non-neuronal membranes. MATERIALS AND METHODS Pre/mration of mouse cerebral cortical astrocyt¢,s Primaw cultures of cerebral cortical astrocytes were prepared from Swiss Albino newborn mice (I-2 days old) by a modification of the method described by McCarthy and de Vellis~'~. Forebrains were removed aseptically from the skulls, the meninges were excised carefully, and the neocortex dissected, The cells were dissociated by passage through needles of decreasing gauges (I.2×40 ram, 0,8×40 mm and 0.5 × 16 ram) with a 5 ml syringe. No trypsin was used for dissociation. The cells were seeded at a density of 10"S/cm-" on

24-well plates in DMEM (Dulbecco's modified Eagle's medium) containing I(}% fetal calf serum (FCS), in a final volume of 500 #l per well, and incubated at 37°C in an atmosphere containing 5% CO: at a 95% humidity. The culture medium was renewed after 4 days of seeding and subsequently twice per week. These conditions yield astrocyte cultures containing 85-90% glial fibrillary acidic protein (GFAP)-positive cells4"s. For dibutyryl cyclic AMP (dBcAMP) experiments, cultures were grown for an additional 2 days in the presence of i mM dBcAMP to induce further differentiation, Secondary cultures were prepared by trypsinizing confluent primary cultures with 0.25% trypsin for 5 rain at 37°C in Dulbecco's phosphate-buffered saline (PBS). The cells were then centrifuged at 1,000x g for 5 rain, resuspended in DMEM containing 10% FCS and finally seeded at a density of 5x 104/cm 2 on 24-weU plates. After 3-5 days, secondary cultures were usually confluent. Binding assay of M-/12~i]VIP to cultures of mouse cerebral cortical astrocytes

Mouse cerebral cortical astrocytes (20-40 p,g of protein per well) were incubated for 2 h (except for kinetic experiments in which incubation time was varied) at 22°C in 500 /zl DMEM containing 10% FCS in the presence of M-[z:'~I]VIP (corresponding to a final concentration of 37 and 100 pM for experiments performed on primary and secondary cultures, respectively) and of various concentrations of unlabeled VIP, PHI or secretin. Following incubation with M-[l:sl]VIP. the medium was aspirated and the free M-[z:'Si]VIP was removed by washing twice with 1 ml DMEM. Astrocytes were finally harvested after brief sonication in I ml DMEM and bound M-[ i-"sl]VlP assessed in a gamma counter. Non-specific binding was determined in each experiment as the amount of total M-[I"'Sl]VIP bound to cultured astrocytes in the presence of I p,M unlabeled VIP. Specific binding was obtained by subtraction of non-specific from total binding, and represented 70-80% and 80-90% of total binding in primary and secondary cultures, respectively.

Preparation oJ' microv,,ssel,~ The microvess¢ls were prepared according to the method described by Iluang and Drummond I'~ with some minor modifications. Briefly, hruins from 12 Swiss male alhino mice (2-3 months old) were removed afld placed on ice, The cerebral cortices were then dissooted, frc~d of plal membranes, surface vessels and white matter, and then homol~efll~edin IS ml of Kr~hs~RInger hicarhoaat¢ halter, ptl 7,4 (KRG), containing 5 mM glucose and I~/~, (v/v) newborn calf serum (buffer At, using a Potter teflon-glass homogenizer (12 up ;tad down str~kcs of the pestle), The homogcnate was then centrifuged at 8iX1× g for 10 rain, The resulting loose pellet consisted of two layers, After removal of the supernutz~nt, the upper layer was diluted in a small volume of butter A and removed, The lower layer, made up of incompletely disrupted tissue, was then homogenized in $ vols, of fresh buffer A and centrifuged as above. The resulting upper layer was added to the previous one, The remaining bottom layer was resuspended in 3 vols, of buffer A, homogenized and combined to the particulate suspension from the first two centrifugations. The mixture was further disrupted by homogenization, centrifuged at I,II{Ht× g for l0 rain and the resulting pellet diluted in 10 vols. of KRG pH 7,4 containing 25% (v/v) newborn calf serum. The homogenate was centrifuged at 3,000× g for 15 rain and the pellet containing microvessels, nuclei and erythrocytes resuspended in 10 ml of KRG pH 7,4 containing 25% (v/v) newborn calf serum. The suspension was then centrifuged at 3,0(H)x g for 15 rain and the pellet homogenized in 2 ml of buffer A, In order to remove nuclei and ¢rythrocytes, the suspension was passed through a glass bead column (0,45-0,50 mm glass bead diameter) and washed with KRO until the effluent was free of nuclei and erythrocytes. Finally, the glass beads were extruded and the entrapped microvessels released by stirring the beads repeatedly with 4 ml of 25 mM Tris-HCI pH ?A containing 2 mM MgCI, and 1 mM ~-mercaptoethanol. The protein concentration determined by the method of Bradford 2 was approximately 150 p,g/ml. Binding experiments were always performed the same day as the microvessel preparation.

Binding assay of M-[ IZ.sl]ViP to cortical microl'essels 120 p.I of microvessel preparation (approximately 18 p.g of protein) were incubated for 1 h (except [or kinetic experiments in which incubation time was varied) at 22°C with 40 p.I M-[l'~'~l]VlP (approximately 17,000 cpm corresponding to a final concentration of 21 pM, dilution performed in 25 mM Tris-HCI buffer pH 7.4 containing 2 mM MgCI2 and I mM /3-mercaptoethanol (buffer B) containing 1% (w/v) bovine serum albumin in th~ p.res,;:.'~ceof various concentrations of unlabeled VIP, PHI or secretin, r-ollowing incubation with M-[IZ~l]'v'lP, the reaction mixture was transferred onto 300/~l of ice-cold buffer B containing 0.32 M sucrose ano 2% bovine serum albumin, and the tubes were then centrifuged at 9,980× g for 2 rain. The supernatants were aspirated and the pellets, containing the radioactivity bound to the microvessels, assessed in a gamma counter. Non-specific binding was determined in each experiment as the amount of total M-[Z2sl]VIP bound to the microvessels in the presence of I p.M unlabeled VIP. Specific binding was obtained by subtraction of non-specific binding from total binding, and represented approximately 50% of total binding.

Preparation of synaptosoma! membrmzesfrom mouse cerebralcortex Synaptosomes were prepared from the cerebral cortex of 2- to 3-month-old mice, according to D o d d e t al s. The cerebral cortices were rapidly dissected and placed in 10 vols. (w/v) of ice-cold 0.32 M sucrose. Usually 8-10 cerebral cortices (1200-1500 rag) were pooled. The tissue was homogenized using a Potter teflon-glass homogenizer (motor speed 800 rpm, 12 up and down strokes of the pestle). The initial homogenate was centrifuged at 830× g for 15 rain at 4°C. The supernatants were divided into aliquots and layered directly onto 4 ml of 1,2 M sucrose; the volume was brought to 10 ml/tube with 0,32 M sucrose and the tubes centrifuged at 180,000× g for 20 rain at 4°C on a Centrikon T-2070 (Kontron, Ultracentrifuge). The interface between the two sucrose concentrations (containing myelin and synaptosomes) was diluted in 0.16 M sucrose, layered onto 4 ml of 0,8 M sucrose and centrifuged at 180,0()0× g for 20 rain at 4°C. The supernatant was discarded and the pellet containing the synaptosomes resuspended, at a protein concentratioq of approximately 2 mg/ml, in buffer B (see binding assay to cortical microvessels) and stored at -8()°C4-~. The protein concentration was determined by the method of l~wry ''b using bovine serum albumin as a standard.

Binding a,~sayof M.I I'~IIVIP to ,~'yneq~temmlalmo~ah~nes The procedure used was that described by Staun-Olsen el al. 4' with minor modifications. Isolated syn|tptosomes were diluted to a concentration of I mg/ml with buffer B. 120 #l of synaptosomes (approximately 120/zg of protein) were incubated for I h (except for kinetic experiments in which incubation time was varied) at 37°C with 40 p,I of M-[l'~Si]VIP [16,0(10 cpm corresponding to a final concentration of 19,5 pM, dilution performed in buffer B containing I% (w/v) bovine serum albumin] in the presence of various concentrations of unlabeled VIP, PHI or secretin (40 #l). Following incubation with M-[12"~I]VIP, the reaction mixture was transferred onto 300 ~l of ice-cold buffer B containing 0.32 M sucrose and 2% bovine ~rom albumin, and the tubes centrifuged at 9,980x g for 2 rain. The supernatants were aspirated and the pellets, containing M-[12"~I]VIP bound to synaptosomes, gently washed with 500/zl of ice-cold buffer B containing 1% bovine serum albumin. After aspiration of this buffer, the radioactivity bound to the tissue was assessed in a gamma counter. Non-specific binding was determined in each experiment as the amount of total M-[12"~I]VIP bound to the synaptosomes in the presence of 1 /~M unlabeled VIP. Specific binding was obtained by subtraction of non-specific binding from the total binding, and represented approximately 50% of total binding.

cAMP assay For cortical microvessels and astrocyte cultures, cAMP levels were measured according to the following procedure. A microvessel suspension (10/zg of protein) or astrocytes in 35 mm diameter Petri dishes (120-150 p.g of protein) were incubated for 20 min at 37°C in KRG buffer pH 7.4 with isobutylmethylxanthine (IBMX; final con-

centration I mM) and VIP or secretin. The reaction was stopped by a brief sonication, the tissue was then boiled for 10 rain at 95°C and centrifuged for 2 rain at 9,980 x g. An aliquot of the supernatant was taken to assess cAMP levels by radioimmunoassay using [l~Sl]cAMP as tracer (Amersham). Protein content was measured using the method of Bradford 2.

Glycogen assay The procedure used was that described by Sorg and Magistretti 41. Briefl}', after 14-21 days in culture, astrocytes were used for the glycogen assay. Culture medium was removed and the cells were incubated for 4 h in a low glucose (5 raM) serum-free DMEM. This medium was used to dissolve the peptides, which were added for 30 min in the same conditions (37°C). The reaction was stopped by washing the cells with ice-cold PBS, and by adding 400 p.! of 30 mM HCI. The cells were sonicated and the suspension was used to measure glycogen as described by Nahorsky and Rogers 32. Briefly, three 100 p.I aliquots were sampled. In the first aliquot, 300 p.I of acetate buffer were added (0.1 M, pH 4.65). in the second, 300/~1 of a ~lution containing 10% of amyloglucosidase (10 mg/ml) in acetate buffer were added and the mixture was incubated at room temperature for 30 rain. After incubation with amyloglucosidase, 2 ml of Tris.HCI buffer (0.1 M, pH 8.1) containing MgCl~ (3.3 raM), ATP (0.2 raM), NADP (25 ~g/ml), hcxokinase (4 p.g/ml) and glucose-6phosphate dehydrogenase (2 p.g/ml) were added, and the mixture was incubated at room temperature for 30 rain. The first aliquot was treated identically. The fluorescence of the NADPH formed was then read in a Perkin.EImer fluorometer (excitation: 340 rim; emission:.450 rim). The first aliquot gives the sum of glucose and glucose-6-phosphate, while the second gives the sum of glycogen, glucose and glucose-6-phosphate; the amount of glycogen is then determined by the difference between the first two aliquots. The third aliquot was used to ineasure the protein content with the method of Bradford :. M.[I~"~I]VIP was purchased from Amersham, the peptides were from Bachern (Bubendorf/Swilzerland) and all other reagents from Sigma Chemical Co. (St. Louis, MO).

RESULTS

Astrocyte cultures

In a first series of experiments the association and dissociation kinetics of M-[1251JVlP binding to intact cultured astrocytes were examined. As shown in Fig, I A, specific M-[12'~l]VIP binding increased with incubation time, reaching a maximal steady level between 120 and 180 rain. The calculated association constant k+ I was 1.483 × 10~min- i. M- i (inset Fig. 1A). The dissociation curve was rnonophasic, virtually all M-[12"~I]VIP specifically bound being displaced by 1 /~M unlabeled VIP within 90 rnin (Fig. IB). The calculated dissociation constant k _ I w a s 0.0173 rain -I (inset Fig. IB). Competition curves for VIP and the structurally-related peptides PHI and secretin are shown in Fig. 2. These experiments indicated that while VIP and PHI effectively competed for binding sites labeled by M[1251]VIP, secretin was without effect, in fact, an unexpected but consistent observation was a moderate ( ~ 25%) increase in M-[t2Sl]VIP binding in the presence of 1 p,M secretin. Scatchard analysis of the competition curve of VIP indicated a single class of binding sites with a Ko of 3.3 nM and a Bin.,,x of 565 fmol/mg

lO

.f A

Q

-~ ~

i'

~ ¢-)

I(]0

...*

• "

7s

~2 -g '° •

~• . . . . . . "

"~ s

~ ' "%.

2

pMbound

\ 0 0

!

I

I

I

I

25

50

7S

100

125

25

a 150

175

~

- ~

200

& I

Time (min.)

0

'

-I0

'

-9

.,8

-1'

-6

Log [Peptide] (M)

"

.8

6

•

~

"\

O5 O0

Fig. 2. Competition of VIP (e). PHI (A) and secretin ( l ) , with M-[J:sl]VIP for specific binding sites in primary cultures of mouse cerebral cortical astrocytes. Primary astrocytes were incubated for 2 h with 37 pM M-[12sl]VIP in the presence of increasing concentrations of unlabeled peptides. Absolute values of total and non-specific binding were (in cpm) 857+ 12 (n = 4) and 223± 18 (n = 3), respectively. Results, which correspond to the mean of triplicate determinations, are plotted as percentages of M-[zzsI]VIP specifically bound against concentration of unlabeled peptide. S.E.M. were less than 10% of the means, Inset shows a Scatchard representation of the VIP competition curve.

*

~ "°r

"~. 1

4

B

l l n l e ([llill,)

3

'-;"

1 0

I

I

I

I

20

40

60

80

100

~nl~ (nlin,)

Fig, I, Time course for the association ;rod dissocia:io. ,)f M-II'slJVIP binding to primary cultures of mouse cerebral cortic.'d dstrogyies, A: time cour),e of association. Primary astrocytes were incubated for various periods of time with 37 pM M.[I"~I]VIP, in the presence (non.specific binding) or absence (total hinding) of Ip.M unlabeled VIP. Specific Binding was defined hy suhtractin8 non.specific from total binding, Results, which correspond It) the mean of triplicate determinations, are expressed i. fmol/m~ protein S,E.M, wt:~e less than IIF~; el' the means, Inset shows a linear representation of the data from wllich the asst~iation rat~ constant ( k , i ) was determined, B: lime course of dissociation, Cultured astrocytes were incubated in the presence of 37 pM M.II-'~I]VIP. After 2 h of incubation. I #M

enzyme activity or receptor expression. For this reason, astrocyte cultures were pretreated with 1 mM dBcAMP for 48 h prior to establishing competition curves for VIE Scatehard analysis indicated no significant effect of dBcAMP treatment since a single class of binding sices was revealed with properties similar to those observed in untreated cells, i.e. K a of 3.9 nM 3000

2SO0

unlabeled VIP was added and radioactivity bound to the astrocytes

assessed at various dissociation times. Results, which correspond to the mean of triplicate determinations, are expressed in final/me protein S.E,M, were less than 10~ of the means. Inset shows a linear representation of the data from which the dissociation rate constant ( k . t) was determined,

protein (inset Fig. 2). The K~ for PHI was i0 nM. As an index of functionality for these VIP binding sites, the effect of VIP on cAMP formation was examined. As indicated in Fig. 3, VIP stimulates cAMP formation in a concentration-dependent manner with an EC~0 of 7 nM. A number of reports 14'-': have indicated that treatment of cultured astrocytes with the membrane-permeable analog of cAMP, dibutyryl cAMP (dBcAMP), favors the differentiation of these cells as indicated by various parameters such as morphological appearance,

2OOO

(~c5

Ja~ o

_[

s~

0

h a m l " 4),0

41,1

ALO Log

.I',!

-TO

.i~S

.4LO

[VIPI ( M )

Fig, 3, Concentratlon-re~ponse curve of VIP on cAMP accumulation in primary astrocyte cultures, Primary astrocyte cultures were incubated for 20 rain in the presence of I raM IBMX and of increasing concentrations of VIP, Following incubation, the reaction was stopped as described under Materials and Methods and cAMP levels assessed by radioimmunoassay. Results, which correspond to the mean of triplicate determinations, are expressed in pmol/mg protein. Basal cAMP levels were 68 + 2 praol/rag protein (n = 3).

and Bmax of 742 fmol/mg protein (Fig. 4). Treatment with dBcAMP did not affect the protein content of the cultures. Passaging of astrocyte cultures from primary to secondary plating (see Materials and Methods) is an occasionally used procedure, reported to increase cellular homogeneity. Trypsinization of the primary cultures is necessary for dissociation and replating of the cells. Since trypsin treatment has been reported to influence the binding characteristics of VIP to intestinal epithelial cells and to rat brain membranes "~3"4°,we examined the binding characteristics of M-[=251]VIP binding in passaged astrocyte cultures. Association and dissociation kinetics of M-[u2"~I]VIP in passaged astrocyte cultures are shown in Fig. 5A and 5B, respectively. Association of M-[nz'~I]VIP was timedependent, reaching maximal steady levels b.etween 100 and 180 rain, i.e., a time-course similar to that observed in primary cultures (Fig. 1A). Dissociation was also similar to that observed in primary cultures, with virtually all specific binding being displaced within 90 m'in. (Fig. 5B). However competition curves with VIP and PHI yielded results markedly different from those observed in primary cultures. Thus, a curvilinear Scatchard plot was obtained from VIP competition experiments (Fig.

120

too .

,

'~%t

~

go

0 ,t,,

10

gO ~O pM bound

40

2,5 Q J~

2.0

~,.

1.5

A ~-0.5

0.0

i

i

i

i

|

2s

so

;.s

;no

~.s

i

i

~so ~;'s

200

Time (min.)

3.0

== ,-,

j

'.°

n

.~

~ ~.

,.0 0,5

0,0 0

i 20

i 40

i 60

i 80

100

Time (rain,) Fig. 5. Time course for the :msociation and dissociation of M-[ J2"~I]VIP binding to secondary cullures of mouse cerebral cortical astrocytes, A: time course of :~ssociation, Secondary astrocytes were incubated for various periods of time with 100 pM M-[u2'~I]VIP,in the presence (non-specific binding) or absence (total binding) of I #M unlabeled VIP, Specific binding was defined by subtracting non-specific from total binding, Results, which correspond to the mean of quadruplicate determinations, ar~ expressed in fmol/mg protein S.E.M, were less than I()% of th~ means, B: time course of dismlciation. Sec. ondary astrocytes were incubated in the presence of 100 pM M. [l:sl]VIP, After 2 h of incuh,tion, I /~M unlabeled VIP was added and radioactivity hound to the astrocylcs assessed at various dissocia. tion limes. Results, which ,re the mean of quadruplicate dclerminatkms, are expressed in fmol/mg protein S,E,M, were less th.n I l l ~ of the mcans.

:>, m

o

•1 0

.9

..0

-?

Log [VIP] (M) Fig. 4. Effect of dBcAMP pretreatment on M-[12'~I]VIP bindi.g to primary cultures of mouse cerebral cortical astrocytes. Primary astrocytes were prepared as described under Materials and Methods. After reaching confluence, astrocytes were incubated for 2 days in the presence ([]) or absence (e) of I mM dBcAMP. Cultures were incubated for 2 h with 37 pM M-[i2"~I]VIP in the presence of increasing concentrations of unlabeled VIP. Absolute values of total and non-specific binding were (in cpm) 900+27 (n = 4) and 219+ I (n = 3) for dBcAMP.treated astrocytes. Results, which correspond to the mean of triplicate determinations, are plotted as percentages of M-[k?'~I]VIP specifically bound against concentration of unlabeled VIP. S.E.M. were le~ than 10% of the means, Inset shows a Seatchard representation of the VIP competition curves.

6), which was resolved for two classes of binding sites 3`n3.3s with Kus of 3.6 and 41.3 nM and Bin. x of 356 and 881 fmol/mg protein respectively. Furthermore, the IC.~,) for PHI was approximately 20 times higher, with a value of 220 riM. (Fig. 6). However, as in primary cultures, secretin did not compete for sites labeled by.M-[n2"~l]VIP(Fig. 6). Secretin has been shown to share with VIP a number of actions in the CNS, such as for example the stimulation of cAMP formation ? and of glycogenolysis 24. In view of the observed inability of secretin to compete for sites labeled by M-[n251]VIP, the possibility was considered that secretin could exert its actions by binding to receptors independent from those for VIP. For this reason we examined the effects of se-

cretin on glycogen hydrolysis in primary cultures of mouse cerebral cortical astrocytes. As shown in Fig. 7A, secretin promotes a concentration-dependent glycogenolysis, with an ECs0 of 0.5 riM. For comparison the ECso of PHI to elicit this action is 6 nM (Fig. 7B). Similar results were observed in secondary cultures that had been subjected to trypsinization and replating (not shown). Cortical microt'essels and synaptosomes The binding characteristics of M-[~esl]VIP to mouse cerebral cortical astrocytes were compared to those observed in isolated intraparenchymal microvessels (Fig. 8) and in synaptosomes. Like astrocytes, microvessels and synaptosomes were prepared from mouse cerebral cortex. As shown in Fig. 9A, M-["sl]VIP binds to isolated microvessels in a time-dependent manner. Specific binding reaches maximal values within 60 min and remains stable at least up to 100 rain. As shown in Fig. 9B, dissociation of M-[~"~I]VIP is also time-dependent. Competition curves for VlP and the structurally related peptides PHI and secretin are shown in Fig. 10. A curvilinear Scatchard plot was obtained from the competition curve with unlabeled VIP, indicating the presence of two classes of binding sites with Kas of 1.4 and 30.3 nM and Bm,~ of 7.1 p m o l / m g protein and 73.8 pmol/mg protein, respectively x'~''~. Furthermore,

t30 a

100

~j,

. , , . ' t . . . . . . '~,, " ~

~

P" "~,

mo,im,ms'ma'O

20

o 15 v

A

10

£: g

$

E

•

.

• 10

°

.

-9

,

.

41

,

.

-7

, -6

log [SecretinI ( M )

es

E

B

!

so

=

B

dl

, • !0

_

, .9

.

i

_

,45

J .r

.

, .6

Io8 [PHil (MI Fig. 7. Concentration.response curves of the glycosenolytic effect of secretin (A) and PHI (B) in primary cultures of mouse cerebral cortical astrocytes. Astn~cyteswere incubated as indicated in Materials and Methods, Olycogen levels were measured by fluorometry as previously described~:,~i, Results, which correspond to the mean of triplicate determinations, are expressedin nmoi/;ng plx)t of glycogen hydrolyzed, Basal levels of glycogen corresponded to 28:1:3 nmol/mB protein OI - 3) and 32 ± 2 nmol/m8 protein (n - 3) for secretin and PHI, respectively.

m m

so

~,,

4o

,°

"..

,

\

? i "'5 °''., • 11

-10

\

, -g

41

-7

-6

Log [Peptide] (M) Fig. ~. Competition of VIP (el, PHI (,,) and secretin ([3), with M-[Iz~IIVIP for specific binding sites in secondary cultures of mouse cerebral corlical astrocytes, Secondary astrocytes were incubated for 2 h with II}0 pM M-[iZsl]ViP in the presence of various concentrations of unlabeled peptides, Absolute values of total and non-specific binding were (in cpm) 1864:1:103 (n~4) and ~ 6 : t l i ( n - 3 ) , respectively. Results, which corresl~md to the mean of quadruplicate determinations, are plotted as percentages of M-[I~slIVIP specifically bound against concentration of unlabeled Peptide, S,E,M, were less Ihan l i l ~ of the means, Inset shows a Scatchard representation of the VIP competition curve,

similar to what is observed in cultured astrocytes, secretin interacts only marginally and at very high concentrations with M-[~Zsl]VlP for specific binding sites. The ICs0 of PHI in this cell type is 50 nM (Fig. 10). in intraparenchymal microvessels also, VIP stimulates in a concentration.dependent manner cAMP formation, with an ECso of 18 nM (Fig. 11). However, in marked contrast to what is observed in cultured astrocytes, where secretin stimulates cAMP formation 4,4s and glycogenolysis (Fig. 7A) through independent recdptors, in intraparenchymal microvessels, secretin does not stimulate cAMP formation (not shown). A potential glycogenolytic action could not be examined in microvessels, in view of the extremely low glycogen levels detected in this cell type. The association curve for M-[12sl]VIP specific binding to mouse cerebral cortical synaptosomes is similar to that observed for microvessels, with equilibrium

Fig. 8. Photomicrograph of typical inlraparenchymal microvessels of mouse cerebral cortex. Microvessels were prepared as described under Materials and Methods. Bar ffi 2(} ~m.

being reached within 60 rain. (Fig. 12A). Dissociation is also time-dependent with virtually all M-[Iz~I]VIP specifically bound being displaced by 1 /zM unlabeled VIP within 120 min (Fig. 12B). Competition curves for VIP, PHi and seeretin are shown in Fig. 13. In agreement with previous reports in rat and guinea pig cerebral cortex a~'aT'42-44, a curvilinear Scatchard plot of the competition curve with VIP was obtained, revealing the presence of two classes of binding sites with Kds of 4.9 and 42.8 nM and Bmax of 316 and 1578 fmol/mg protein, respectively "~'13'as. Synaptosomal membranes are the only preparation examined in which secretin competes for sites labeled by M-[I'sl]VIP, with an ICs0 of 150 nM (Fig. 13); the ICso for PHI is 35 nM (Fig.

13). DISCUSSION In the present study we have characterized the kinetic and pharmacological properties of VIP binding sites in three distinct cellular preparations of the mouse neocortex. From this analysis, evidence for the existence of three subtypes, with different cellular Iocalizations has emerged. The properties of the three VIP receptor subtypes are summarized in Table I. The first subtype (VIP 1) is ubiquitous and of high affinity with K d S of 3.3 and 3.6 nM (primary and secondary astrocytes, respectively), 1.4 nM (microvessels) and 4.9 nM

(synaptosomes). Secretin does not interact with this site. This latter pharmacological property is clearly indicated by the following observations: (i) secretin does not compete for sites labeled by M-[~Zsl]VIP in primary cultures of astrocytes, a cell type that expresses only high affinity VIP binding sites and (ii) the ICs0 of secretin in synaptosomal membranes (the only preparation in which secretin interacts with VIP sites) is 150 nM (this study and ref, 42), i,e, a value compatible with an interaction with low affinity VIP binding sites. The second receptor subtype (VIP 2) is exclusively present on synaptosomal membranes. It is a low affinity site, with a Kd of 42.8 nM. Seeretin interacts with this site (see above and Fig. 13). The third subtype (VIP 3), is also of low affinity, with a K d of 30.3 nM in intraparenchymal microvessels, and 41.3 nM in secondary astrocyte cultures, it can be distinguished from the other low affinity site (VIP 2) on the basis of two properties. First, secretin does not compete with M[12-~I]VIP for the VIP 3 binding site; second, it is exclusively present on non-neuronal membranes, namely those of microvessels (Fig. 10) and of astrocytes that have undergone trypsin treatment and have been replated as secondary cultures (Fig. 6). in contrast to secretin, PHI, another peptide structurally related to VIP, does not discriminate between neuronal and non-neuronal sites, since it competes with M-[~25I]VIP in astrocytes and microvessels as well as in synaptosomes (Figs. 2, 6, 10 and 13).

VIP receptors on astroeytes

IO0

The presence of functional VIP receptors on astrocytes could be inferred from observations reporting VIP stimulation of cAMP formation in cultured glioblasts 4s and in primary astrocyte cultures (ref. 4 and Fig. 3). Furthermore the presence of VIP binding sites on cultured astrocytes was revealed by in vitro autoradiography TM. One of the potential functional consequences of cAMP formation in astrocytes was demonstrated to be the stimulation of glycogenolysis-'-~. The ECs.s for VIP in promoting cAMP formation and glycogenolysis are 7 nM (Fig. 3) and 3 nM (ref. 41). respectively, which are values that correlate remarkably well with the Kd of VIP 1 receptors on astrocytes, i.e. 3.3 nM (Fig. 2).

4O

A

c: ..~ o ..D

80

"~

so

tJ

~

4o

"

20

=r, pM Bound

o • 10

-9

-8

-7

-6

Log IPeptide] (M) Fig. 10. Competition of VIP (e). PHI (&) and secretin (D), with M-[t:~I]VIP for specific recognition sites in intr:tparenchymal microvessels. Cortical microvesscls were incubated for I h with 21 pM M-[I-"~I]VIP in the presence of increasing concentrations of unlabeled peptides. Absolute values of total and non-specific binding of M-[I~"~I]VIP to microvessels were (in cpm) 2125_+28 ( n = 3 ) a n d 113[)+58 (n = 3), respectively. Results, which correspond to the means of triplic;ite determinations, are plotted as percentages of M.[I"'~IJVIP specifically hound against concentration of unlabeled peptide. S.E,M. were less than 10% of the means. Inset shows a Scatchard representation of the VIP competition curve.

__.'~ to i

o

tS

10

4S

eO

75

=.z.__,__--=~=

10S

90

lhtt~ (ttti.,) 4O

=~

Exposure of astrocyte cultures to dibutyryl cAMP has been shown to trigger a number of processes that lead to marked changes in cell structure and function. For example, under the influence of dBcAMP, astrocytes convert from a fiat, epitheloid appearance charactcristic of poorly differentiated cells to adopt the morphology of 'mature' astrocytes, i.e. star-shaped cells with a small perikaryon from which emerge numerous

~o

BOO

40O

;g

,~

1o

300

m

o 200

o

o

30

z

i

60

90

. . . .

~__ 120

~. Th|le

100

{Inln.)

Fig. 9, Time course fi)r the association and dissociation of M-[ ~"'~I]VIP binding to intraparenchymal micmvcsscls. A" lime ~=,{mrsetff association. Microvessels were incubated f,r vari,us periods of time with 21 "',~ M.[~'~I]VIP in the presence (non-specific binding)or absence (total bi~ding) uf I #M unlabeled VIP, Specific binding was defined by subtracting non.specific from total binding. Results, which correspond to the mean of duplicate determinations, are cxpre~,~cd in fm~d/mg protein, B: time course or dissociation, Microvessels were incubated in the presence of 21 pM M-[~z~IJVIP, After I h incubation. I ,aM unlabeled VIP was added anti radioactivity bound to the microvessels a~,~essed at various di~m~ciati~m times, Results, which o.q|'~JSl~.utd to the mean of duplicate determinations, are expressed in fmnl/mg protein,

0

I

10 `9

10,¢

10-7

10 4

Lo~ [VlPl (M) Fig. I1. Concentration-response curve of VIP on cAMP accumulation in intraparenchymal microvessels, Cortical microvessels were incubated for 20 rain in the presence of I ~:~M IBMX and of increasing concentrations of VIP. FoUowing incubation, the reaction was stopped as de~ribed under Materials and Methods and cAMP levels ussessed by radioimmunoassay, Results, which correspond to the mean of triplicate determinations, are expressed in pmol/mg protein. Basal cAMP levels were 16a± 16 pmol/mg protein (n = 3).

so

.g=

110

:

/

40

A

'~ 0

\

rJ

°°~1

lO

i 30

i

o

\

IO0

lS

i

i

i

45

60

75

90

"°

:

,o

t

:

Time (rain.)

I

" .

: .%:::0

0

2S0

.

500

"\ "" " , , " - . \. , .

o

.\

750 t000

•

pM Bound 0

=

-11

-10

-9

-8

-7

-6

Log [Peptide] (M) Fig. 13. Competition of VIP (e), Pttl (A) and sccretin (ra), with M-[Iz~I]VIP for specific recognition sites in synaptosomal membranes. Synaptosomes were incubated for I h with 19.5 pM M[ta~l]VIP in the presence of increasing concentrations of unlabeled peptides. Absolute values of total and non-specific binding were (in cpm) 14t2 + 7 (n = 4) and 715 + 5 (n = 4), respectively. Results, which correspond to the mean of triplicate determinations, are plotted as percentages of M-[i2"~I]VIP specifically hound against concentration of unlabeled peptide. S.E.M. were less than 10% of the means. Inset shows a Scatchard representation of the VIP competition curve.

0.0

' 30

60

90

J

120

Thne (rain.) Fig. 12, Time course for the association and dissociation of M[t:'~l]V1P binding to synaptosomal membranes. A: time course of association. Synaptosomal membranes were incubated for various periods of time with 19.5 pM M.[12'~I]VIP in the presence (non-

specific binding) or absence (total bindinl~) of I /zM unlabeled VIP, Specific hindin~ was defined by subtracting non-specific from total blndinu, Results, whicll correspond to the mean of quintuplicate determinations, are plotted as percentages of M-[="'~I]VIP specifically bound, S,E,M, were less than I()r/;, of the means, B: time course of dissociation, Synuplosomal membranes were incubated i. the presence of Ig,5 pM M-[t'~I]VIP, After ! h incuhutian, I ~M unlabeled VIP was added and radioactivity bound to synaptosomes assessed at various dissociation times. Results, which correspond to the

meansof quintuplicate determinations, are expressed in fmol/n.g protein S,E,M. were lessthan 10% of the means.

processes 22'-~n. Monoamine receptors and the signal transduction cascades to which they are coupled are also influenced by dBcAMP pretreatment =4. Recently we have observed that, when astrocyte cultures are treated for 48 h with dBcAMP and then allowed to recover for 4 h, a lO-fold increase in glycogen levels is observed, when compared to control cultures 4~. In view of these marked effects, the concept has been entertained, but not yet established, that dBeAMP may promote the differentiation of astrocytes j4. From the results reported in Fig. 4 it is clear that dBcAMP treatment does not influence the properties of VIP receptors in astrocytes.

However, dissociation of primary astrocyte cultures by trypsin treatment followed by replating as secondary cultures, another manipulation which is often used to prepare cultures of neural tissues, influence the properties of VIP binding sites. Thus, a low affinity site, for which secretin does not compete, is expressed in astrocytes submitted to such treatment (Fig. 6). The molecular mechanisms that underlie this process remain unclear: unmasking of a recognition site by limited proteolysis could for example be hypothesized. One fact is however firmly established by these experiments: the low affinity site that can be unmasked in astrocytes is of the non-neuronal subtype (VIP 3), since similarly to the low affinity VIP site on microvessels (Fig. 10) but unlike the low affinity site on neuronal membranes (Fig. 13), it is insensitive to seeretin. Regarding secretin, it is clear that it can trigger cellular processes in astrocytes by acting on receptors TABLE I

Propm'ed class!fication oJ' VIP receptor subt)7;es in cerebra/cortex Subtype

Localization

VIPI

astrocytes (I, i l ) " microvessels synaptosomes

K,I hrM) 3.3; 3.6 1.4

Competition by st'crelin

4.9

-

VIP 2

synaptosomes

42,8

+

VIP 3

astrocytes (II)" microvessels

41.3 30.3

-

" l, primary cultures; II, secondary cultures (see 'Materials and Methods').

10 that are distinct from those for VIP. As discussed earlier, secretin does not compete for sites labeled by M-[I~I]VIP, even when a low affinity site is unmasked (Fig. 6). However, secretin promotes glycogenolysis with an EC~ of 0.5 nM (Fig. 7A). This value is in striking agreement with the K d of 0.2 nM reported for the binding of secretin to brain membranes, i.e. a preparation that contains both neuronal and non-neuronal membranes =a. It is therefore possible that the careful analysis reported by Fr6meau et al. ~°, primarily reflects the binding of secretin to astrocyte membranes, it should be noted here that on the basis of elegant pharmacological experiments in which adenylate cyclase activation by VIP and secretin was analyzed, Chneiweiss et al. "~ hypothesized that VIP and secretin had independent receptors on striatal astrocytes. Results reported in the present article verify this hypothesis. VIP receptors on intraparenchyma/ microcessels The presence of functional receptors for VIP has been previously demonstrated in bovine and porcine cerebral arteries 4". Thus, in bovine middle cerebral artery, Scatchard analysis of VIP binding data revealed the presence of two recognition sites with Kds of 0.2 and 11 nM, respectively. Occupation of these receptors by VIP resulted in stimulation of cAMP formation and

va~orelaxation '¢'~7'4¢',Secretin did not interact with sites labeled by M-[I"'~I]VIP and was devoid of biological activity ~r'. Similar results were observed in the present report which was carried out on intraparenchymal microvessels rather than on large cerebral arteries. Thus two recognition sites were revealed, with Kds of 1,4 and 30.3 nM, respectively; furthermore, in keeping with what is observed in large cerebral arteries, secretin interacted only marginally at very high concentrations with VIP receptors. The marginal interaction of secretin is likely to be due to contamination by neuronal membranes. Furthermore, secretin did not stimulate cAMP formation (not shown). This latter finding indicates that, in contrast to what is observed in astrocytes, functional secretin receptors are absent from cercbrovascular cells. Receptors for VIP on intraparenchymal microvessels are functional, as indicated by the fact that VIP stimulates cAMP formation in these cells, with an EC~. of 18 nM (Fig, II), lntraparenchymal microvessels appear to be particularly enriched in VIP binding sites, as the observed receptor density is in the pmol/mg of protein range, While VIP receptors on large cerebral arteries are likely to be activated by VIP released from fibers having their origin in extracerebral circuits, e.g,, sphenopalatine ganglion =2, the presence of functional

VIP receptors on intraparenchymal microvessels strongly supports the view that the VIP-containing intracortical bipolar neurons 23"31 may exert local regulatory functions on blood flow or vascular permeability within the cerebral cortex. VIP receptors on synaptosomes Although the binding characteristics of M-[~I]VIP to synaptosomal membranes have already been described in guinea pig and rat cerebral cortex 36'37"42-44, they were nevertheless analyzed in the present study in order to compare under similar experimental conditions, VIP receptors in various cells types in the same species (mouse) and brain area (cerebral cortex). As previously reported for guinea pig and rat, the binding of M-[1~'51]VIP to mouse cerebral cortical synaptosomes reveals the presence of two sites with KdS of 4.9 and 42.8 nM, respectively, with secretin interacting with the low affinity site (Fig. 13). This latter observation clearly indicates that secretin represents a useful pharmacological tool to distinguish low affinity neuronal VIP receptors from other subtypes. In addition to the presence of ~/IP receptors on synaptosomal membranes, structural as well as functional evidences support the notion that VIP can affect neuronal activity. First, axe-dendritic VIP-positive synapses have been identified on the shafts of small-diameter dendrites in the noocortex'~4; second, VIP alters in a slowly-reversible manner the excitability of identified cortical neurons ~, Results reported in this article indicate that VIP receptor subtypes arc differentially-expressed in astro. cytcs, microvessels and in synaptosomal membranes, thus supporting the view that VIP, released from VIPcontaining intracortical neurons, can affect the activity of three distinct cell types, The interaction with high affinity VIP receptors (VIP 1) on astrocytes results, among other possible effects, in the stimulation of cAMP and glycogenolysis, the latter process contribut. ing to the maintenance of local energy metabolism homeostasis ''~, The interaction with receptors on intraparenchymal microvessels (VIP 1 a n d / o r VIP 3) leads to cAMP formation; the functional consequence of the occupation by VIP of receptors on the microvasculature remains to be determined. Local blood flow and permeability regulation are likely candidates 23. Activation of VIP receptors (VIP ! and VIP 2) on neurons results in the modulation of excitability~. After this article was accepted for publication, lshihara et aL (Neuron, 8 (1992) 811-819) have reported the isolation, from a rat lung eDNA library, of a functional eDNA clone encoding the VIP receptor. VIP receptor mRNA is expressed in various rat tissues

II

including the brain, where highest levels are detected in the cerebral cortex and hippocampus. Furthermore, in addition to the 5.3 kb mRNA found in all tissues examined, the brain contains two other mRNAs species (2.5 and 1.3 kb). This observation lends further support to the existence of at least three VIP receptor subtypes in the brain. Acknowledgements, This work was supported by a grant from Fends National Suisse de la Recherche Scientifique (31-26427.89) to PJ.M. The authors wish to thank Mrs. M. Emch for expert secretarial help, Ms, Maillard and M, Gasser for technical assistance. REFERENCES I Besson, J., Dussaillant, M., Marie, J.-C., Rost~ne, W. and Rosselin, G., in vitro autoradiographic localization of vasoactive intestinal peptide (VIP) binding sites in the rat central nervous system, Peptides, 5 0984) 339-340, 2 Bradford, M,M., A rapid and sensitive method for the quantitation of microgram quantities of protein using the principle of protein-dye binding, Ann. Biochem., 72 (1976) 248-254. 3 Chamness. G.C. and McGuire, W.L., Scatchard plots: common errors in correction and interpretation, SteroMs, 26 (1975) 538542, 4 Chneiweiss, H., GIowinski, J, and Pr~mont, J., Vasoactive intestinal polypeptide receptors linked to an adenylate .~yclase, and their relationship with biogenic amine- and somatostatin-sensitive adenylate cyclases on central neuronal and glial cells in primary cultures, J. Neurochem,, 44 (1985) 779-786. 5 Chneiweiss, H., Glowinski, J, and Pr6mont, J., Do secretin and vasoactive intestinal peptide have independent receptors on striatal neurons and glial cells in primary cultures?, d. Neumchem., 41 (1986) 608-613. 6 Couvineau, A,, Gammeltoft, S. and Laburthe, M., Molecular characteristics and peptidc specificity of v:tsoactive intestinal peptide receptors from rat cerebral cortex, J. NeurtMwm., 47 (1986) 1469-1475. 7 Deschodt-Lanckman, M., Rohherecht, P. and Christophe, J., Characterizatkm of VlP-sensitiw adenylate cyelase in guinea-pig brain, FEBS Lett., 83 (1977) 76=80, 8 Dodd, P,R., Ilardy, J,A., Oakley, A,E., Edwardson. J,A,, Perry, E,K, lind Dclaunoy, J.P,, A rapid method for Prel)nring synapto. sprees: comparison with alternative procedures, Brain Rcs., 226 (1981) 1()7-118, 9 Ferron, A., Siggins, G.R, and Bloom, F,E,, Vasoactive intestinal polypeptide acts synergistically with norepinephrino to depress spontaneous discharge rate in cerebral cortical neurons. Proc. Natl, Acad, &'i. USA, 82 (1985) 8810-8812. 10 Fr~meau, R.T., Jr., Jensen, R.T., Charlton, CG., Miller, R.L., O'Donhoe, L. and Moody, T,W., Secrelin: specific binding to rat membranes, J. Neurosci., 3 (1983) 1620-1625. II Gray, E.G. and Whittaker, B.P., The isolation of nerve endings from brain: an electron microscopic study of the cell fragments of homogeffation and centrifugation. J. Anat., 96 0962) 79-88. 12 Hara, H., Hamil, G.S, and Jacohowitz, D.M., Origin of choliner. gic nerves to the rat major cerebral arteries: coexistence with vasoactive intestinal polypeptide, Brain Res'. Bull., 14 (1985) 179188. 13 Hart, H.E., Determination of equilibrium const:mts and maximum binding capacities in complex in vitro systems. I. The mammillary system, Bull. Math. Biophy~., 27 (1965) 87-98. 14 Hertz. L., Dibutyryl cyclic AMP treatment of ,'tstrocytcs in primary cultures as a substitute for normal morphogenic and ffunctiogenic' transmitter signals. In J.M. Lauder tEd.). Moh,cular Aspects of Det'elopment and Aging of the Nert'ous System. Plenum, New York, 1990, pp. 227-242. 15 Huang. M, and Drummond, G.I., Adenylate cyclase in cerebral microvessels: action of guanine nucleotides, adenosine and other agonists, Mol. Pharmacol,, 16 (1979) 462-472.

16 Huang, M. and Rorstad, O.P., Effects of vasoactive intestinal polypeplide, monoamines, prostaglandins and 2-chloroadenosine on adenylate cyclase in rat cerebral microvessels, J. Neurochem., 40 (1983) 719-726, 17 Huang, M. and Rorstad, O.P., Cerebral vascular adenylate cyclase: evidence for coupling to receptors for vasoactive intestinal peptide and parathyroid hormone, J. Neurochem., 43 (1984) 849856. 18 H6sli, E. and H6sli, L., Autoradiographic localization of binding sites for vasoactive intestinal peptide and angiotensin 11 on neurons and astrocytes of cultured rat central nervous system, Neuroscience, 31 (1989) 463-470, 19 Lor~n, I., Emson, P.C., Fahrenkrug, ,L, Bj6rklund, A., Ahmets, J., Hakanson, R. and Sundler, F,, Distribution of vasoactive intestinal polypeptide in the rat and mouse brain, Neuroscience, 4 (1979) 1953-1976. 20 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J,, Protein measurements with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 21 Magistretti, PJ., VIP neurons in the cerebral cortex. Trends Plmnnacol. $ci., II 0990)250-254. 22 Magistretti, P.J., Manthorpe, M., Bloom, F.E. and Varon. S.. Functional receptors for vasoactive intestimfl polypeptide in cultured astroglia from neonatal rat brain, Regtd. Pcpt., 6 (1983) 71 - 80. 23 Magistretti, P.J. and Morrison, J.H., Noradrenaline-and vasoactire intestinal peptide-containing neuronal systems in neocortex: functional convergence with contrasting morphology, Neurosck'ncc, 24 (1988) 367-378. 24 Magistretti, P.J., Morrison, J.H., Shoemaker, W.J., Sapin, V. and Bloom, F.E., Vasoactive intestinal polypeptide induces glycogenolysis in mouse cortical slices: a possible regulatory mechanism for the local control of energy metabolism, Prec. Natl. Acad. SoL USA, 78 (1981) 6535-6539. 25 Magistretti. P.J. :ind Schorderet, M., VIP ,'rod noradrenaline act synergistically to increase cyclic AMP in cerebral cortex, Natttre, 308 (1984) 280-282. 26 Martin, J.L., Dietl, M.M, Hof, P.R., Palacios, J.M. and Magistretli, P,J,, Autoradiographic mapping of [mono[I;Sl]iodo. Tyre°,MelOIV].vasoactive inlestinal peptide binding sites in the rat brain, Neumscience, 23 (1987) 539=505, 27 Martin, J,L, and Magistrcltl, PJ., Pharmacological evidence for a role of voltage.sensitive Ca" +-channels of the T type in the release of vasoactive intestinal pepli(,lc evoked by K ' ii') mouse cerebral cortical slices, Neuro,wh,m'(,, 30(1989)423=431, 28 Martin, J.L. and Magistretti, P,J,, Release of vasoactive intestinal peptide in mouse cerebral corlex: evidence for a rol~ of arachidonic acid mctabolites, I Neumsci,, 9 (1989) 2536=2542, 29 McCarthy, K.D. trod de Vellis, J. Preparation of separalc as. troglial and oligodendroglial cell cultures from rat cerebral tissue, J, Cell Biol., 85 (1980) 890-9112. 30 Moonen, G,, Heine;t, E. and Goesscns, G., Comp:mltivc ultra. structural study of the effects of serum-free medium and dibu. tyryl-cyclie AMP on newborn ;'tit astroblasts, Ce/I Tissue Res,. If~7 (1976) 221-227, 31 Morrison, J,H., Magistretti. P,J.. Benoh, R. and Bloom, F.E,. The distribution and morphological characteristics of the inlracortical VIP-positivc cells: an immunohistochemical analysis, Btw#a Res,, 292 (1984) 269-283, 32 Nahorski, S.R, and Rogers, KJ.. An enzymatic fluorometric micromethod for determination of glycogen. Amd. Biochem,. 49 (1972) 492-497, 33 Ogawa, N,, Mizuno, S,, Mort, A., Nukina, I, and Yanaihara. N,. Properties and distribution of vasoactive intestinal polypeplidc receptors in the rat brain, PeptMes, 6, Suppl, I (1985) 103-1(}9, 34 Peters, A. and Harriman, K,M.. Enigm:ltic bipolar cells of rqt visual cortex, Jr. Comp. Nem'o/., 267 (1988) 409-432, .45 Ouik, M,, Iversen, L.L. and Bloom, S,R,. Effect of vasoactive intestinal pcptide (VIP) and other peptides on cAMP accumulation in rat brain, Biochcm. PharmacoL, 27 (1978) 2209-2213. 36 Rohberecht. P., De Neef, P.. Lammens, M.. Deschodt-Lanckman, M. and Christophe. J., Specific binding of vasoactive intestinal

12 peptide to brain membranes from the guinea pig, Eur. J. Biochem., 90 (1978) 147-154. 37 Robberecht, P,, Konig, W., Deschodt-Lanckman, M,, De Neef, P. and Christophe, J., Specificity of receptors to vasoactive intestinal peptide in guinea pig brain, Life Sei., 25 (1979) 879-884. 38 Rosenthal, H.E., A graphic method for the determination and presentation of binding parameters in a complex system, Anal Biochem., 20 (1967) 525-532. 39 Said, S.i. and Mutt, V., Polypeptide with broad biological activity: isolation from small intestine, Science, 169 (1970) 1217-1218. 40 Sarrieau, A., Laburthe, M. and Rosselin, G., Intestinal VIP receptors: differential effect of trypsin on the high and low affinity binding sites, Mol. Cell. Endocrinol., 31 (1983) 301-313. 41 Sorg, O. and Magistretti, PJ., Characterization of the glycogenolysis elicited by vasoactive intestinal peptide, noradrenaline and adenosine in primary cultures of mouse cerebral cortical astrocytes, Brain Res., 563 (1991) 227-233. 42 Staun-Olsen, P., Otlesen, B., Barrels, P.D., Nielsen, M.H., Gammeltoft, S. and Fahrenkrug, J., Receptors for vasoactive intestinal polypeptide on isolated synaptosomes from rat cerebral cortex. Heterogeneity of binding and desensitization of receptors, 3'. Nt'~rochem., 39 (1982) 1242-1251.

43 Staun-Olsen, P., Ottesen, B., Gammeltofl, S. and Fahrenkrug, J., The regional distribution of receptors for vasoactive intestinal polypeptide (VIP) in the rat central nervous system, Brain Res., 330 (1985) 317-321. 44 Staun-Olsen, P., Ottesen, B., Gammeltoft, S. and Fabrenkrug, J., VIP binding sites on synaptosomes from rat cerebral cortex: structure-binding relationship, Peptides, 7, Suppl. 1 (1986) 181186. 45 Stoyanov, T., Martin, J.L. and Magistretti, P.J. VIP binding sites in primary cultures of astrocytes, Fur..I. NeuroscL, SI (1988) 111. 46 Suzuki, Y., McMaster, D., Huang, M., Lederis, K. and Rorstad, O.P., Characterization of functional receptors for vasoactive intestinal peptide in bovine cerebral arteries, J. Neurochem., 45 (1985) 890-899. 47 Taylor, D.P. and Pert, C.B., Vasoactive intestinal polypeptide: specific binding to rat brain membranes, Proc. Natl. Acad. Sci. USA, 76 (1979) 660-664. 48 Van Calker, D., Miiller, M. and Hamprecht, B., Regulation by secretin, vasoactive intestinal peptide and somatostatin of cyclic AMP accumulation in cultured brain cells, Proc. Natl. Acad, Sci. USA, 77 (1980) 6907-691 I.