SYNAPSE 8~144-151(1991)

High Affinity Histamine Binding Site Is the H9 ReceDtor: Characterization and Autoridiogriphic Localization in Rat Brain PAUL CUMMING, CHRISTOPHER SHAW, AND STEVEN R. VINCENT Kinsmen Laboratory of Neurological Research, Department of Psychiatry (P.C., S.R.V.) and Ophthalmology (C.S.), University of British Columbia, Vancouver, B.C., Canada, V6T 1W5

KEY WORDS

Histamine, Binding, Rat, Basal ganglia, Brain

ABSTRACT The binding of the histamine autoreceptor (H3)agonist [3H]-N"-methylhistamine ([,H]-N-MeHA) was examined in 25 pm thick rat forebrain sections. The specific binding was saturable and of high affinity: Scatchard analysis indicated a & of 2 nM and a B, of 25 f 3 fmolhection. Under similar conditions, [3H]-histamine ([3H]-HA) bound with a & of 8 nM and a B, of 20 2 2 fmol/section. Competition studies indicated that both ligands bound an identical site which had the pharmacologxal characteristics of the H3binding site. The high affinity binding of [3H-N-MeHAwas sensitive to the presence of 5'-guanylyl-imidodiphosphate,indicating that the binding site is likely coupled to a G-protein. Autoradiographic studies indicated the E3H]-N-MeHAbinding to be greatest in the nucleus accumbens, striatum, substantia nigra pars reticulata, and certain cortical areas. Striatal quinolinic acid lesions greatly reduced binding in both the striatum and ipsilateral substantia nigra, while 6-hydroxydopamine lesions of the nigrostriatal dopamine system were without effect on binding. Therefore, most of the H, binding sites in the basal ganglia are on striatonigral projection neurons. Cortical quinolinic acid lesions greatly reduced H, binding in cortex, indicating that the binding in cortex, as in striatum, is largely on intrinsic neurons, rather than on afferents such as histamine nerve terminals.

INTRODUCTION Histamine is able to interact with nervous tissue through at least three pharmacologically distinct binding sites. The H2 receptor stimulates the formation of cyclic AMP (Hegstrand et al., 1976); the HI receptor potentiates the effect of H2 agonists on cyclic AMP formation (Palacios et al., 1978a) and stimulates the hydrolysis of phosphotidylinositol in brain tissue (Fisher and Agranoff, 1987).The H3 receptor, which is pharmacologically distinct from the preceding classes, has been described as a presynaptic site controlling the synthesis and release of histamine (Arrang et al., 1983).Using the potassium-evoked release of [,H]-histamine as an index of autoreceptor activation, N"-methylhistamine was characterized as a prototypic H, agonist (Arrang et al., 1983,1990; van der Wolf et al., 1987), while thioperamide was described as an antagonist (Arranget al., 1987). If the H, binding site serves principally as an autoreceptor, its distribution in brain would be expected to correlate highly with the density of histaminergic innervation. In the present study we characterize the binding properties of the H, agonist [3Hl-N"-methylhistamine ([,H]-N-MeHA) and compare it to tritiated histamine (C3H]-HA)binding in rat brain sections. The two ligands 0 1991 WILEY-LISS, INC.

were found to bind with high affinity to an identical site: the high affinity histamine binding site. Lesion studies indicate that the majority of this H3 binding is associated with forebrain neurons rather than with afferent histaminergic terminals. MATERIALS AND METHODS Adult male Wistar rats (250-300 g) were killed by cervical dislocation and the brain quickly removed, blocked, and frozen. For binding studies, 25 pm thick cryostat sections, each weighing 2.0 +- 0.1 mg, were prepared in the coronal plane. Sections were taken at the level of the greatest dimension of the striatum. Fresh sections were mounted on chrom-alum coated glass slides and stored at -70°C for no more than 1 week. Sections were incubated for 45 minutes at room temperature in 150 mM sodium phosphate buffer (pH 7.5) containing 100 pM dithiothreitol (Sigma), 2 mM MgC12, and varying concentrations of the radioligands. [3H]-HA(32 CYmmol, New England Nuclear) was included at 0.4-80 nM and L3H1-N-MeHA (80 Ci/mmol, New England Nuclear) was included at 0.15-25 nM. Received September 28,1990; accepted in revised form December 12,1990.

HISTAMINE RECEPTORS IN THE BRAIN

Nonspecific binding was determined by the addition of 5 pM of the chiral H3 agonist R-(a)-methylhistamine oxalate (a-MeHA, Research Biochemicals). Sections were washed three times in 4°C buffer (20 sec), dried under an air line, and scraped from the slides with a razor blade. The radioactivity in each section was determined by liquid scintillation counting (Packard Tri-carb 4530) after the addition of 5 ml aqueous cocktail (ACS, Amersham). The & and B,, were determined by Scatchard analysis (Munson and Rodbard, 1980). To examine possible regulation of binding by guanylyl nucleotides, rat forebrain sections from three rats were incubated in the presence of 100 pM 5'-guanylyl-imidodiphosphate (GppNHp) in addition to E3H1-N-MeHA at concentrations from 1 to 500 nM. At ligand concentrations between 50 and 500 nM, specific activity was reduced by the addition of unlabelled N-MeHA (Calbiochem). Non-specific binding was determined by the addition of 10 mM a-MeHA in these experiments. In competition studies, radioligands were included at their respective &s as determined above: 8 nM in the case of L3H]-HA and 2 nM in the case of [3H]-N-MeHA. Competition curves were produced by the addition of competitors at various final concentrations ranging from to M. The competitors used were histamine hydrochloride, mepyramine, cimetidine, and sulpiride (all from Sigma), .r-methylhistamine (Calbiochem),impromidine (Smith, Kline, and French), harmaline hydrochloride (Aldrich), trazadone (Bristol), desmethylimipramine (Merle Dow), haloperidol (McNeil Labs), thioperamide (gft of Dr. J.-C. Schwartz), and phencyclidinehydrochloride (gift of Dr. C. Blaha). In some animals neurochemical lesions were produced with either quinolinic acid (&A) or 6-hydroxydopamine (6-OHDA)in order to determine the anatomical localization of histamine binding. Three male Wistar rats (250-300 g) were placed in a stereotaxic apparatus while under pentobarbital anesthesia (60 mgkp i.p.). Using the coordinates of Paxinos and Watson (1982),we placed a syringe in either the sensorimo-0.26, DV: -2.5-0.5, ML: -3.4) tor cerebral cortex (AP: or the dorsal striatum (AP: -0.26, DV: -5.5, ML: +3.4) and 1 p1 of 300 mM &A (Sigma, pH 8) was infused. QA-lesioned animals were sacrificed 4 days after surgery, at which time neuronal destruction is complete in the zone of the lesion while axon terminals from outside the lesion remain largely intact (Schwarcz et al., 1983; Shaw et al., 1989). Three other animals received 1 p1 of 18 mM 6-OHDA (Sigma) in ascorbic acid solution directed towards the left medial forebrain bundle (AP: +4.5, ML: 1-15, D V +1.5) and were sacrificed 7 days after the 6-OHDA lesions, at which time presynaptic dopamine receptors are lost (Feuerstein et al., 1981). Portions of striatum from the 6-OHDA lesioned animals were taken for determination of catecholamine content by reversed phase HPLC with electrochemical detection as described previously (Cumming et al., 1987).

145

In the lesion studies, 25 pm cryostat sections were prepared at the level of striatum and mesencephalon. Slides containing four sections each were incubated ir, L3H1-N-MeHAat concentrations of 0.5,1,2,5,and 12 nM and washed as described above. The saturation binding curves in forebrain hemisections from some 6-OHDAlesioned animals were determined directly by liquid scintillation counting. Other slides were exposed to tritium-sensitive film (Amersham Hyperfilm) for a 6 week period after which time the films were conventionally developed and fixed. Regional optical densities were quantified by autoradiographic image analysis (Imaging Research Inc., St. Catharines, Ontario) employing calibrated standards (Amersham). RESULTS Scatchard analysis of [3H]-HA binding in rat forebrain (Fig. 1)yielded a Kd of 8 nM and B,, of 20 I 2 fmolhection, while Scatchard analysis of l3H1-N-MeHA binding (Fig. 2) yielded a & of 2 nM and B, of 25 2 3 fmollsection. Hill coefficientsfor the two ligands did not differ significantly from unity (Fig. 3). Addition of 100 pM GppNHp greatly interfered with specific [3H]-N-Me-HAbinding: 94% of the high affinity binding disappeared with the residual component having an apparent 3 nM affinity. At ligand concentrations above 25 nM, a new binding site appeared with & of approximately 200 nM and B, of approximately 50 fmolhection. Unfavorable signal-to-noise ratio at high ligand concentrations made it difficult to determine these quantities with precision. In the competition studies, nonspecific binding in forebrain sections was 12% of the total at 2nM L3H1N-MeHA and 25% of the total at 8 nM L3H1-HA.Ic507sfor the competition of [3H]-N-MeHAbinding by some histamine ligands (Fig. 4A) and by various other psychoactive drugs were estimated from the interpolation of the curves at half-maximal specific binding after subtraction of the nonspecific binding. The most potent competitors of [3H]-N-MeHA binding (Fig. 4) were the H3 agonist a-MeHA (IC,o = 0.4 nM) and the H3 antagonist thioperamide (ICs0 = 3 nM). Impromidine, an H3 antagonist and H, agonist, had an ICs0 of 79 nM, while burimamide, an H3 and Ha antagonist, was a relatively more potent competitor of [3H]-N-MeHA (IC50= 20 nM). .r-Methylhistamine, the ring-methylated histamine metabolite, had little ability to compete with 13H]-N-MeHA.IC50)sfor a variety of psychoactive drugs, determined by interpolation were as follows: haloperido1 (25 pM), sulpiride (45 pM), trazodone (45 pM), phencyclidine (70 pM), desmethylimipramine (DMI, 100 pM), and harmaline (>lo0 pM). The competition by some drugs for [3H]-HAbinding was also examined and is illustrated in Figure 4,B. Autoradiographic images of [3H]-N-MeHAbinding in rat brain indicated that the highest binding was in the nucleus accumbens, striatum and substantia nigra pars

146 3

2 (D

0 F

X

I

--. U

m

1

P. CUMMING ET AL.

12

7

9-

*

0 7

10

10

60

80

Ligand Concentration InM)

['HI-histamine

x

6 -

v \

m

D

12 0 1

0

5

10

15

20

25

0

\

,

I

10

20

\

30

Ligand Bound (fmole/section)

Ligand Bound (frnole/section)

Fig. 1. Scatchard analysis of [3H]-HA binding in rat forebrain sections. Each point represents the mean of four determinations which differedby less than 10%.Inset: The correspondingsaturation binding curves.

Fig. 2. Scatchard analysis of L3H1-N-MeHAbinding in rat forebrain sections.Each point represents the mean of four determinations which differed by less than 10%.Inset: The correspondingsaturation binding curves.

reticulata. Dense binding was also noted in the globus pallidus, amygdala and pyriform cortex. In the cerebral cortex, there was considerable heterogeneity in the maximal binding. Maximal binding in the anterior cingulate cortex was 30% higher than in the adjacent sensorimotor cortex (Table I; see also Fig. 5A), but binding was very low in the retrosplenial cingulate cortex (Fig. 5B). Binding in the insular cortex appeared to be almost equal to that in the basal ganglia. Binding was heterogeneous within the somatosensory cortex (Fig. 5C, arrow), possibly related to the barrel field. Within cortex, binding had a laminar distribution, with relatively higher binding in layers 1 , 3 and 5 (Fig. 5C). However, there was little specific binding in layer 3 of the primary visual cortex (Fig. 5B). There was relatively little binding in the dorsal hippocampus and moderate binding in the dentate gyrus of the ventral hippocampus (Fig. 5B). Non-specific binding was very low in the mesencephalon (Fig. 5F) and all other areas examined. Scatchard analysis of L3H]-N-MeHA binding in the striatum after a 6-OHDA lesion to the left medial forebrain bundle did not indicate any effect of this treatment (Fig. 6A). HPLC analysis indicated a 95% depletion of striatal dopamine in the 6-OHDA lesioned striata (control levels = 65 nmoVg tissue). In three separate determinations of binding to forebrain sections, the dopaminergic lesions were without significant effect on binding parameters (Table I). In the autoradiographic study, the unilateral 6-OHDA-lesion had no effect on the binding in either striatum (Fig. 5A) or substantia nigra (Fig. 5B). A representative Scatchard plot of [3H]-N-MeHA binding in striatum after a unilateral striatal QA-lesion indicates a major reduction in maximal binding with no apparent changes in affinity (Fig. 6B). Striatal QA-

1 I

1 -

E

m

I

-m \

0-

- -SO -

slope = 0.91

0

-1.5

/

I

I

1

-10

-9

-a

-7

log,, [Ligandl Fig. 3. Hill plots of the binding of 13H]-N-MeHAand I3H1-HA to cryostat sections of rat forebrain.

lesions produced great reductions in binding in both striatum (Fig. 5C) and in the substantia nigra ipsilatera1 to the lesion (Fig. 5D). The treatment produced a 70% decrease in striatal B,, without significantly altering the apparent affinity (Table I).Nissl-staining of the QA-lesioned forebrain section (Fig. 5E) showed the tissue to be intact, but devoid of neurons in the region of the lesion. The effect of the QA-lesion on binding of [3H]-NMeHA in cerebral cortex is illustrated in Figure 7. The unilateral QA-lesion in sensorimotor cortex produced a 70% decrease in maximal binding without altering the affinity (Table I). The binding in cingulate cortex ipsilateral to the lesion was not altered with respect to the contralateral side. The autoradiographic appearance of

HISTAMINE RECEPTORS IN THE BFWN

-

60 -

0-1, -11

OJ,

-11

,

,

,

-10

-9

-8

I

-10

,

-9

I

-8

, -7

I

-7

,

,

-6

-5

I

-6

I

-5

~, -4

-3

I