Quantitative light microscopic autoradiographic localization of α2-adrenoceptors in the human brain

Brain Research, 585 (1992) i 16-127 ~'~ 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00

116

BRES 17818

Quantitative light microscopic autoradiographic localization of a2-adrenoceptors in the human brain Julio Pascual a,b, C a r m e n del Arco a, Antonio M. Gonzfilez a and Angel Pazos

a

Department of Physiology and Pharmacology, Unit of Pharmacology and t, Sen'ice of Neurology, University Hospital "Marquds de Valdecilla", Faculty of Medicine, Unicersity of Cantabria, Santander (Spain) (Accepted 28 January 1992)

Kc~. words: a:-Adrenoceptor; Bromoxidine; Human brain; Receptor autoradiography

in the present work the anatomical distribution of a2-adrenoceptors in the human central nervous system was studied in detail by quantitative autoradiography using the selective a 2 agonist ['~H]bromoxidine ([3H]UK-14304) as a ligand. Only postmortem tissues from subjects free of neurological disorders were used in this study. Very high or high densities of a2-adrenoceptors were found along layers i and ill in non-visual neocortex, layers Ill and IVc of the visual cortex, CAt field - - stratum lacunosum-moleculare - - and dentate gyms - - stratum granularis - - at the hippocampal formation, nucleus arcuatus at the hypothalamus, locus ceruleus, nucleus dorsalis of vagus and at the stratum granularis of the cerebellar cortex. Relevant densities of a:-adrenoceptors were also observed along the remaining layers of neocortex, nuclei centralis, mediaUs and corticalis at the amygdala, anterior thalamic group and rotundocellularis nuclei, paraventricular and ventromedial hypothalamic nuclei, substantia innominata, superior colliculus -- stratum zonale - - and lateral periaqueductal area at the midbrain, nucleus tractus solitarii and dorsal horn ~ substanti:l gclatinosa - - of the spinal cord. ['~H]Bromoxidine specific binding was very low or negligible in the remaining brain areas. Although a general parallelism between the distribution of these receptors could be observed for the rat and human brain, dramatic species differences in the level of a~-receptors were found in several brain areas, such as thalamus, amygdala or cerebollar cortex, in general, the distribution of a,-adrenoceptors in the human brain found here was parallel to that described for the noradrcnergic presynaptic terminals in the mammalian central nervous system, lending som~ weight to the proposed predominant presynaptic localization of these receptors, The relevance of the anatomical distribution of a,-adrenoceptors in the human brain for a better knowledge of the neurochemistry of neuropsychiatric disorders is discussed.

INTRODUCTION Half a century ago, Ahlquist classified adrenergic receptors in two types, a and /3I. More recently, the existence of two subtypes for both the a (al and a2) and the//(/31 and /3,) types has been described 2~,4s. Furthermore, the properties and anatomical distribution of u2.receptors have already been extensively studied by ligand binding procedures in the rat central nervous system ( C N S ) 3'4~'51. The biochemical properties and the pharmacological profile of human brain a2-reccptors have been well studied by means of binding methods in membrane homogenates '~,1o,27,st,5-~,On the other hand, very recently knowledge of the molecular biology of the a2.adrenoceptor has become available (see ref. 17). in contrast to that, the information

about their precise regional localization is still limited. By using [~H]p-aminoclonidine and autoradiographic procedures, a2.adrenoceptors have been visualized in tissue sections of the human brainstem 42, By comparing the results obtained in several human forebrain areas both in membrane homogcnates and tissue sections with the a2-agonist [3H]bromoxidine (['~H]UK14304)~, we have reported that [3H]bromoxidine is a feasible ligand to localize and quantify a,-receptors in human postmortem tissue sections by autoradiographic techniques "~', However, a detailed microscopic mapping of these sites is still lacking. Moreover, a better knowledge of the distribution of a2-receptors in a large sample of human brains appears as very necessary to study the putative changes that could take place in these receptors in some neuropsychiatric conditions

Correspondence: A. Pazos, Dept. of Physiology and Pharmacology, Unit of Pharmacology, Faculty of Medicine, University of Cantabria, C/Cardenal Herrera Oria s/n, 39011 Santander, Spain. Fax: (34) 42-347942.

117 TABLE I

General characteristics of the brain tissue Case

Age (years)

Postmortem delay O~ours)

1

F

69

3

2 3 4 5 6 7 8 9 10

M M M F F F M F M

57 73 57 84 61 78 64 73 63

16 il 8 7 5 8 11 21 18

I1

F

73

7

12 13 14 15 16 17 18 19 2O 21

M F M M F M M M F M M 9F/13M

81 61 19 49 19 52 39 46 66 45 41 57_+17* 19-84

22

Totals Range

Sex

22

5 14 20 17 32 16 42 17 32 20 39 17+!1 * 3-42

* Mean + S.D.

such as Alzheimer's disease 4-~''t2'46, different parkinsonisms t 1,34 or depression 2¢'. in the present paper, we show the complete mapping of the a2-receptors in the whole CNS. Our data extend and confirm Probst et al.'s brainstem mapping 42 as well as our very preliminary results in the human forebrain .a~'. MATERIALS AND METHODS Twenty-two brains from subjects dying without clinical and neuropathological evidence of neuropsychiatric disease were obtained at necropsy from the Service of Pathology, University Hospital 'Marquis de Valdecilla'. To our knowledge none of the patients had received psychotropic medications before death. Table I summarizes the characteristics of these cases with respect to age, sex and postmortem delay. At necropsy, brains were promptly removed and dissected. One hemibrain was fixed in 10% buffered formalin and processed for further routine examination. The remainder of the CNS was kept at 4"C for 30 rain, cut into 4-6 mm thick blocks and stored at -70"C. The blocks were later brought to -20"C and mounted onto microtome chucks. Ten ,¢m tissue sections were cut using a microtomecryostat, mounted onto gelatine-coated glass slides and stored at - 2 0 " C until used. Rat brain tissues (adult Wistar rats, both sexes, body weight 150-350 g) were processed in parallel with the human tissues. a2-Adrenoceptors were labelled with ['~H]bromoxidine (82.7 Ci/mmol, New England Nuclear) as previously described "~'. Briefly, after a 15 rain preincubation at room temperature in Tris-HCI buffer (50 mM, pH 7.7) containing 0.1 mM MnCI2, slide-mounted 10/~m tissue sections were Incubated with 6 nM ['~H]UK-14304 for 90 rain under the same conditions. Following the incubation, sections were washed for 5 rain in ice-cold buffer and then dried in a cold air

stream. Non-specific binding was routinely defined as that remaining in the presence of 10 tzM phentolamine or, alternatively, with I ttM adrenaline. Sections were exposed at 4°C for 10 weeks to tritium-sensitive film (3H-Ultrofilm). The autoradiograms were analyzed and quantified using a computer-assisted image analysis system (Microm-IP, Microm, Barcelona, Spain), Appropriate standards exposed together with the tissues, allowed the transformation of densitometric readings, obtained after several measurements in each nucleus, into receptor densities (fmol/mg protein) 5°. RESULTS

As it has been described in a previous work, binding of [3H]bromoxidine to human brain tissue sections was of high affinity, saturable and with a pharmacological competition profile corresponding to t~2-receptors36. In the present work we have observed a widespread distribution of a2-adrenoceptors throughout the human brain. Important regional variations in the density of a2-receptors were seen over the different areas and nuclei. The [aH]bromoxidine binding to white matter tracts was negligible. The non-specific binding obtained by coincubation with 10/~M phentolamine was rather homogeneous in all brain regions ranging from 0 to 15% of the total binding, except over the substantia nigra where non-specific binding represented more than 80% of the total binding. When 10 -6 M adrmaline was used to define non-specific binding, it ranged from 10% to 25% of the total binding. Tables II-IV summarize the quantitative values of ['~H]bromoxidine binding obtained in different brain structures at a saturating concentration (6 nM).

Variation with age, sex and postmortem delay The densities of a2-adrenoceptors in a particular brain region varied from case to case. In a separate work '~5, we have extensively analyzed the influence of variables such as i~ge, sex and postmortem delay on a2.receptor density. In summary, sex and postmortem delay, In summary, in the range of this study, sex and postmortem delay did not significantly influence the density of a2-adrenoceptors. By contrast, we found age.dependent, regionally specific decreases in the [3H]bromoxidine-specific binding. In general, with the exception of the amygdaloid complex, all the forebrain regions examined showed a striking age-dependent decline in autoradiographic grains, while the infratentorial areas kept the az-adrenoceptors with age 3"~. Regional distribution of t~2-adrenoceptors in the human brain Autoradiographic images shown in Figs. 1-4 illustrate the regional distribution of [-~H]bromoxidine binding in rat and human brain. In the following de-

118 scription, we will consider density values as very high when above 800 fmol/mg protein, high when ranging from 600 to 800 fmol/mg protein, intermediate from 400 to 600 fmol/mg protein, low from 200 to 400 fmol/mg protein and very low those below 200 fmol/mg protein.

Cerebral cortex (Table II, Fig. 1A,B) Relevant densities were observed among the different neocortical areas studied; only the primary visual area (A17) presented levels of binding above the average and a different laminar distribution. The distribution of a2-receptors in human neocortex was laminar,

C

D

II

Fig. I. Autoradiographic images of [~H]bromoxidine binding to human brain cortex, Dark regions represent areas with high densities of binding sites. A: anatomical distribution of ct,-adrenoceptors in frontal cortex demonstrating very high levels of autoradiographic grains along layer I (I) and high levels along layer III (liD, B: a:,adrenoceptors in the calcarine cortex exhibiting very high levels along layer III (III) and high levels along layers IVc (IVc) and I (I), C: hippocampal level showing high densities of aa-receptors along CA t field - - stratum lacunosum-moleculare -- (CA l) and dentate gyrus - - stratum granularis - - (DO), D: amygdalar level presenting low amounts of autoradiographic grains. Note, however, a slight predominance of ['~H]bromoxidine binding over nuclei corticalis (CO), medialis (M) and centralis (CE), as compared to nuclei lateralis (LA), basalis (BA) or basalis accesorious (BAc), Bars -- 2 mm.

119

with important differences among the different layers. The layer Ill in A17 was sharply demarcated, being a cortical area with very high density of a2-receptors. Layer l of the non-visual cortex and layer IVc in the visual cortex showed important densities with their values falling in the high-very high level. Layer III in the non-visual cortex and layer IV in the calcarine cortex presented high densities of autoradiographic grains, while the remaining neocortical layers showed intermediate-low densities.

Hippocampal.formation (Table Ill, Fig. It?) a2-Adrenoceptors also exhibited a laminar distribution in the hippocampus. The densities of these receptors ranged between very high and very low, depending on the layer. The densities measured in the anterior hippocampus were about 20% higher than the posterior levels. The highest densities of a2-receptors in the whole central nervous system were observed along the stratum lacunosum ~ moleculare of the CA I subfield, and in the stratum granularis of the hippocampal gyrus dentatus. The hilus and the stratum lacunosum moleculare at the CA a subfield showed :,,igh amounts of autoradiographic grains. The remaining regions presented low or intermediate densities of these receptors,

TABLE ili Density of a ?-adrenoceptors in hippocampus Brab~ area

n

Specific binding (fmol / m g prot) mean ± S.E.M.

Range

Hippocampus, anterior Subiculum Str. lac.-moleculare Str. radiatum Str. pyramidalis Str. oriens

5 5 5 5

CA 1 field Str. lac-moleculare Str. radiatum Str. pyramidalis-oriens

5 5 4

Dentate gyrus Str. moleculare Str. granularis

Hilus

408± 294± 214± 258±

38 63 64 78

290814823-

521 473 420 563

1373± 359 351 + 104 312+ 53

750-2917 142- 750 130- 391

5

482 ± 114

5 5

1 135 ± 206 713 ± 177

109- 862 604-1950 307-1440

Hippocampus, medial-posterior Subiculum Str. lac-moleculare Str. radiatum Str. pyramidalis

10 9 l0 10

296± 163± 119+ 127+

42 26 17 24

Str. lac-moleculare Str. radiatum Str. pyramidalis

10 10 10

Str. oriens

10

948 ± 237± 196+ 152+

110 25 26 17

463-1437 104- 376 84- 362 75- 229

7 8 8

570± 68 303± 65 277± 52

212- 809 91- 585 122- 605

Str, moleculare Str. granularis

10 10

Hilus

10

254± 49 927 ± 131 458± 71

38- 454 403-1950 146- 883

Str. oriens

125695223-

582 339 215 276

CA I field

CA 2-3 fields

Str. lac-moleculare TABLE !!

Str. pyr-radiatum

Density of a,.adn, noceptors it, neocort~

Str. oriens Dentate gyrus

Brain awa

Frontal cortex Layer i Layer ll Layer !!1

n

Spec(fic binding ([mol/ mg prot ) mean :t: S,E.M.

Range

14

828:1:71

Layers iV-Vl

14 14 14

472:k 40 625 :k 59 315:t: 39

Gyros cinguli Layer i Layer !! Layer II! Layers lV-Vl

5 5 5 5

645± 411 ± 546± 283±

Parietal cortex Layer ! Layer li Layer !!! Layers IV-VI

7 7 7 7

684-1-105 337± 67 539 ± 108 265± 66

262-1 I I0 174- 658 241-1072 83- 537

Temporal cortex Layer I Layer I! Layer III Layers IV-VI

14 14 14 14

681± 366 ± 581± 289 ±

48 27 42 27

314199314167-

998 593 819 492

Basal ganglia and basal forebrain (Table IV, Fig. 2A, C) Very low or low densities were found in caudate-

9 9 9 9 9

411 + 1025 ± 548± 724 + 364±

28 8o 72 78 49

303- 594 642-1369 249- 675 478-1025 184- 618

putamen-accumbens, while the globus pallidus exhibited negligible densities of [:~H]bromoxidine specific binding. The nucleus basalis of Meynert and the claustrum presented the highest labelling in this brain area, with densities in the low range.

Visual cortex Layers I-II Layer I11 Layer IV Layer IVc Layers V-VI

40 49 38 33

382-1292 195- 724 307- 984 44- 537

with very low densities in the external layers of the subieulum in the posterior hippocampus.

486267420194-

Amygdaloid complex (Table IV, Fig. ID)

726 540 640 368

The distribution of autoradiographic grains in the amygdala was rather homogeneous with low or very low densities in most nuclei. Only the nucleus lateralis of the amygdala showed very low densities of autoradiographic grains. The nuclei medialis, corticalis and centralis presented the highest levels of a2-receptors in the amygdaloid complex, all within the low range.

120

Thalamus and hypothalamus (Table V, Fig. 2B)

Midbrain (Table Vi, Fig. 3A)

in general, the thalamus presented a poor concentration of autoradiographic grains. Only the nucleus rotundocellularis showed low but still relevant densities of a2-receptors in this brain area. Thalamic nuclei ventralis anterior, anterior, habenula and medial geniculus presented very low-low densities, whereas the remaining thalamic nuclei exhibited nearly negligible densities in autoradiographic grains. On the other hand, the hypothalamus showed intermediate densities over the nucleus arcuatus and lowintermediate levels on the nucleus paraventricularis and nucleus ventro-medialis, while the remaining hypothalamic nuclei exhibited low or very low densities of a.,-adrenoceptors.

Relevant densities of autoradiographic grains were seen in some midbrain nuclei, such as superior colliculus - - stratum zonale - - and postero-lateral periaqueductal area, with densities in the intermediate level. The remaining layers of the superior colliculus, the oculomotor nucleus, the mesencephalic trigeminal nucleus, the nucleus interpeduncularis and the ventraltegmental area showed low densities of [3H]bromoxidine specific binding, while the remaining midbrain nuclei presented very low-negligible densities.

Pons (Table V1, Fig. 3B) In this brainstem area the locus ceruleus showed relevant, in the intermediate level, densities of ct 2-

P

Fig. 2. Autoradiogranhic images of [~H]bromoxidine binding over several subcortical forebrain areas, A: anterior striatum level showing low densities of c~2-adrenoceptors over caudate (C), putamen (P) and accumbens (ACC.~ nuclei. B: thalamus-hypothalamus level, The thalamus exhibits low but still relevant densities of autoradiographic grains over the nuclei anlerior (ANT), ventralis anterior (VA) and rotundocellularis (RO). The hypothalamus presents important levels of a~-adrenoceptors over nuclei arcuatus (AR), ventromedialis (VM) and posterior (PO) and lower levels over the hypothalamie lateral area (LA). C: posterior striatum level showing relevant densities of ot2-adrenoceptors along the claustrum (CL) and nucleus basalis of Meynert (NB), low levels in the putamen (P) and negligible amounts of autoradiographic grains over the globus pallidus (GP). Bars = 2 ram.

121 TABLE IV

TABLE VI

Density of a z-adrenoceptors in amygdala and basal ganglia

Density of a 2.adrenoceptors in the hindbrain

Brain area

n

Brain area

Amygdaloid complex N. medialis N. centralis N. corticalis N. basalis accesorius N, lateralis N. basalis Intercalated cell masses

9 9 9 9 9 9 7

235 4. 33 230 4. 30 209 4. 26 206 4. 25 91 4.12 1304-16 1345:33

98-355 110-360 98-366 98-332 57-180 73-245 33-316

Basal ganglia Caudate, head Caudate, body Putamen Accumbens G. pallidus, lateralis G. pallidus, medialis Claustrum Substantia innominata

5 14 21 14 13 14 14 11

684- 7 1024-10 117 :t: 11 874.11 144. 6 13 4- 3 257 4. 31 171 + 36

37- 85 36-171 34-3(}0 36-198 0- 59 0- 40 122-475 19-340

Specific binding (fmol / mg prot) mean 4. S.E.M.

Range

adrenoceptors. The remaining pontine nuclei presented very low-negligible levels of autoradiographic grains.

Medulla oblongata (Table Vl, Fig. 3C) This brainstem area showed relevant (intermediatelow) levels of ['~H]bromoxidine specific binding in the TABLE V

Density of a :.adrenoceptors in thalamus and hypothalamus Specific binding (fmot / mg prot) mean + S.E.M. Thalamus N. anterior N. ventralis-anterior N. reticularis N. medialis N. ventralis lateralis N. intralaminaris N. rotundocellularis N. lateralis dorsalis N. lateralis posterior N. ventralis posterior N. centromedianus N. ventralis medialis N. habenularis C. geniculatum mediale C. geniculatum laterale N. pulvinaris Hypothalamus N. supraopticus N. paraventricularis N. ventromedialis N. arcuatus N. tuberalis Area lateralis N. posterior Corpus mamillare

Range

8 3 4 9 4 3 6 5 5 7 6 4 2 5 4 3

177 + 30 191 + 34 40 + 17 94+ 15 51 + 7 28+ 7 268 + 34 115 4. 20 68 + 20 45 + 14 40+ 9 55 + 10 152 + 27 143 + 19 116 + ! 8 85 + 22

90-266 122-266 0- 87 47-189 30- 68 13- 42 113-369 65-180 10-125 10-119 7- 58 34- 85 114-190 98-217 62-145 36-129

3 2 11 8 6 11 8 3

278 + 49 3534. 5 328 4. 39 414+64 206 4. 68 150 + 23 229 + 24 129+ 3

159-338 346-360 83-574 242-712 65-570 45-301 93-309 126-135

Midbrain Superior colliculus Str. zonale Other layers Lateral periaqueductai area Oculomotor nucleus Red nucleus Locus niger Pars compacta Pars reticulata Inferior colliculus Trigeminal nucleus N. interpeduncularis Ventral tegmental area Raph~ nuclei

n

Specific binding (fmol / mg prot) mean + S.E.M.

Range

7 7 10 9 9

393 + 67 202 + 39 339+ 28 239 + 29 33 + 13

100-627 21-351 231-804 129-364 0-100

11 8 3 8 11 4 7

37 + 10 43 + 15 124 + 5 237 + 29 194 + 20 150 + 34 49 + 20

0-107 0-114 116-135 115-375 65-298 74-225 5-127

Pons Locus ceruleus Griseum pontis Raph~ nuclei Motor V nucleus Principal sensory V nucleus Formatio reticularis

4 6 5 3 3 5

416 + 60 9 4. 6 15 + 9 164 + 35 169 + 68 294.15

259-548 0- 42 0- 56 91-258 91-258 2- 95

Medulla oblongata Hypoglossal nucleus Nucleus intercalatus N. dorsalis of vagus N. tractus solitarius N. vestibularis N. cuneatus N. tractus spinalis nervi V N. ambiguus N. olivaris N.arcuatus

4 6 5 6 3 3 6 5 7 5

62 4.17 113 + 14 439 4. 80 190 4. 28 27+ 9 16+ 13 81 + 10 37 + 13 33 + 14 80+ 19

14-101 49-162 207-729 131-249 7- 47 0- 47 44-114 0- 73 10-101 21-127

Cerebellum N.dentatus Stratum granularis Str, Purkinje-moleculare

10 13 14

91 + 15 627 +. 56 352 + 45

26-187 346-930 64-563

Spinal cord Dorsal horn Cell marginatus layer Substantia gelatinosa N. propositus Anterior horn Intermedius horn

7 9 8 7 9

222 + 55 295 + 39 94± 8 1! I + 18 188 + 14

73-31 I 177-587 60-120 25-185 131-269

nucleus dorsalis of vagus and over the nucleus tractus solitarii. The remaining bulbar nuclei were in the very low range.

Spinal cord (Table VI, Fig. 3E) In this brain region the dorsal horn showed the most relevant levels of [aH]bromoxidine-specific binding. In fact, the substantia gelatinosa presented intermediate levels and the cell marginatus layer showed low-intermediate levels of a2-adrenoceptors. The anterior and the intermedius-lateralis horns exhibited densities in the low-very low range.

J

o

~9

-m

Ill

C

. . . :

N. ~

.i:.¸: ~!,'::i~.

~

:

•

.

~:.~.~.

123

Fig. 4. Autoradiographic imagc.s of [3H]bromoxidine bindi.g to rat brain coronal sections: A (striatum level), B (amygdalar level), C (midbrain level) and D (cerebellar level), Note the high densities of aa-receptors found along layer I at the neocortex (I), septum (S), CA1 field - - stratum lacunosum-moleculare - - (CAt), posterolateral (LP) and periventricular (PV) thalamic nuclei, hypothalamus (HY), periaqueductal gray area (PQ) and locus ceruleus (LC), By contrast, low densities of autoradiographic grains can be seen over areas such as caudate-putamen (CP) or cerebellum (CE). Bars ,~ 2 ram,

Cerebellum (Table VI, Fie,. 3D) High densitk~s in autoradiographic grains were observed in this brain region over the stratum granularis at the cerebeUar cortex and intermediate-low densities were found over the remaining layers of the cerebeilar cortex. Cerebellar profound nuclei exhibited very low densities in [3H]bromoxidine-specific binding.

Distribution of az-adrenoceptors in the rat brain (Fig. 4) Using [3H]bromoxidine, the anatomical distribution of o~2-adrenoceptors in the rat brain (numeric data not

shown) did not exhibit any significant difference as compared to that reported by Unnerstall et al using [3H]p-aminoclonidine 4'). In fact, as shown in Fig. 4, relevant densities of a2-receptors in the rat brain were found along layer I at the neocortex, septum, CA~ field stratum lacunosum-moleculare - - at the hippocampus, amygdaloid formation, posterolateral and periventricular thalamic nuclei, hypothalamus, periaqueductal grey area and locus ceruleus. Very low densities of a2-adrenoceptors were seen over other rat brain areas such as the caudate-putamen or cerebellum. -

-

Fig. 3. Autoradiographic images of a2-adrenoceptors over several hindbrain areas. A: midbrain level showing relevant densities of a2-receptors over the lateral periaqueductal area (PQ) and substantia nigra (SN), though in this latter structure more than 80% of binding is non-specific. Also note the low level of a2-receptors seen in the interpeduncular nucleus (IP) and the negligible densities found over the red nucleus (RN). B: pontine level demonstrating relevant densities of [3H]bromoxidine binding concentrated in the locus ceruleus (LC), in contrast to the negligible densities found over the griseum pontis (GP). C: bulbar level showing relevant amounts of autoradiographic grains over the nucleus dorsalis of vagus (X) and nucleus tractus solitarii (TS) and very low levels over the hypoglossal (XII) and olivary (O) nuclei. D: cerebellar level demonstrating high amounts of autoradiographic grains along the granular cell layer (GR) and very low densities over the dentate nucleus (DE). E: spinal cord level exhibiting relevant densities of a2-receptors concentrated along the substantia gelatinosa (SG) at the dorsal horn and low densities over the intermedius (IH) and anterior (AH) horns. Bars = 2 mm.

DISCUSSION The present work shows the anatomical distribution of high-affinity a2-adrenoccptors in the human CNS, providing the first detailed quantitative atlas of the distribution of [3H]bromoxidine binding sites all over the human brain. The densities of a2-adrenoceptors reported here are somewhat higher than those obtained in human brain membrane homogenates by using ['~H]bromoxidine -~7'3t'. In the conditions of measurement used in this work the error introduced by the differential regional 'quenching' must be of small amplitude and does not explain these differences -'3. In general terms, for all receptors the densities measured by microdensitometry are somewhat higher than those determined using membrane binding techniques "~. This fact is probably linked to both a depletion in the number of binding sites during the homogeneization process and the absence of anatomical resolution of the membrane binding methods 43. in addition, the numbers of binding sites reported here are h!gher for most nuclei than those reported in human brainstem sections by using ['~H]paminoclonidine 42. Using human brain membrane homogenates, Meana et al. demonstrated that the full agonist ['~H]bromoxidine labels more binding sites (up to 50%) and with higher affinity (3.fold) than the partial agonist ['~H]clonidine -'7. Although other factors, such as age differences, could theoretically contribute to these differences, it seems that this higher affinity could explain the finding of a higher number of a,binding sites when labelled with ['~H]bromoxidine, than with ['~H]clonidine. The densities of ['~H]bromoxidine. specific binding in the human brain found in this work are higher than those reported by us in a preliminary study using a limited number of cases '~¢'.Very probably, this apparent discrepancy could be explained by the fact that the mean age of the human cases reported here is significantly lower than that of the group used in that preliminary study, as well as by the higher level of receptor occupancy reached in the present report.

Anatomical distribution of tL,.receptors The anatomical distribution of [~H]bromoxidine binding sites in the rat brain shown here is totally coincidental to that reported by using ['~H]p-aminocIonidine 4'j. This again confirms the feasibility of both radioligands to label the a,-adrenoceptor. The discovery of species differences between human and rat brain is of great importance, first because of their intrinsic biological relevance, and second for the study and development of new pharmacological corn-

pounds, suggesting that caution is needed in the extrapolation of experimental data from rat to man 37. The pharmacological characteristics of the a2-adrenoceptor seem to be identical both for human and rat brain membranes. Although some receptors show very similar localization in rat and man, other receptors present dramatic differences in their regional distribution. For example, the human hippocampus is extremely rich in am-adrenoceptors, while in the rat brain the density of a~-adrenoceptors over this structure is very low. On the other hand, the thalamus, which is one of the the regions presenting a very high density of these receptors in the rat brain, contains only intermediate to low levels of acadrenoceptors in the human brain 32. Up to now, the scarce information about the anatomical distribution of a2-adrenoceptors in the human brain has prevented the analysis of possible species differences between the rat and the human brain. By comparing the anatomical distribution and density of a2-receptors in rat and human brain as labelled with ['~H]bromoxidine a general parallelism can be observed, though striking species differences can be appreciated in several brain areas. In agreement with other authors' results using human brain membrane homogenates 9."j.2~.s~.Sa, we found relevant levels of [~H]bromoxidine specific binding in the neocortex, mainly in layers ! and Ill. The rat brain also presents high densities of a2-receptors along the external layers. In the hippocampus, the anatomical distribution of t~:-adrenoceptors in the human brain is close to that of the rat brain. For both species the CA I, stratum lacunosum-moleculare, and the dentate cyrus showed very important receptor densities 4~. The distribution of a2-receptors over the amygdaloid formation is parallel in both species, with the most relevant densities in the centralis, mediaUs and corticalis nuclei. However, there are dramatic differences in the level of st, celtic binding of both species, in fact, while in the rat brain these nuclei show high levels of specific binding 4~, in the human brain they only reach low-very low levels. The distribution of [aH]bromoxidine specific binding sites in the basal ganglia and basal forebrain exhibits a good parallelism for rat and human brain. Both species present negligible levels of binding at the globus pallidus, low levels in the caudate-putamen-accumbens and more relevant densities along the substantia innominata 4~,~. in comparison to the neocortex or the hippocampal formation, the human thalamus shows poor densities of ~2-receptors. Significant, though in the low-very low range, densities are only seen over the anterior group nuclei anterior ventralis, anterior, lateralis dorsalis

125 - - and close to the third ventricle - - nucleus rotundocellularis ~ , the remaining thalamic nuclei presenting negligible or very low densities of a2-receptors. In general, this distribution runs parallel to that described for the rat brain 49, though, as occurred in the amygdalar complex, the relative densities in the human brain are clearly below those reported for the rat brain. In contrast to the differences in the level of specific binding observed in the thalamus, the distribution and densities of a2-receptors are clearly comparable in the hypothalamic nuclei, with relevant densities over the nuclei arcuatus, paraventricularis, ventro-medialis or supraoptic for both species 49. The distribution of a2-receptors in the human brainstem as shown here is close to that of the rat brainstem 49, and identical to that reported by using [3H]p-aminoclonidine in human brainstem sections 42. Both species present relevant densities of autoradiographic grains in the nucleus tractus solitarii, nucleus dorsalis of vagus, locus ceruleus, lateral periaqueductal area or in the stratum zonale of the superior colliculus. In addition, the same distribution and similar levels of binding are seen for both species in the spinal cord where the substantia gelatinosa at the dorsal horn exhibits relevant densities of ~2-adrenoceptors for both species, followed by the intermedius and the anterior horns with densities in the low range. In contrast to this close parallelism observed in the brainstem and spinal cord, important species differences c~n be appreciated in the cerebellum, in this brain region only low levels of binding can be observed along the stratum granularis at the cerebellar cortex in the rat brain both by using [3H]p-aminoclonidine and ['~H]bromoxidine. In contrast, in the human brain, the cerebellar cortex exhibits relevant densities of autoradiographic grains with high levels - - similar to those seen in some forebrain cortical areas - - along the stratum granularis and intermediate-low levels in the remaining cerebellar cortical layers.

t~z.Receptors and noradrenergic innervation It is not known the precise synaptic localization of brain a2-adrenoceptors. Some classical pharmacological studies 14'u'47's4 and, more recently, radioligand binding studies have identified a population of presynaptic t~2-receptors ~3'3°'5~. On the other hand, certain central pharmacological actions of a2-agonists and the anatomical distribution of these receptors in some discrete brain areas, such as the rat cerebellum, seem to be better explained by a postsynaptic localization ~s'2~. Using light microscopic autoradiography, the resolution is insufficient to distinguish pre- vs post-synaptic localization. As reviewed by Unnerstall et al., in the rat

brain, with very few exceptions (most dramatic being the cerebellar cortex), regions which have a2-binding sites are innervated by noradrenergic or adrenergic neurons 4'~. Although the distribution of noradrenergic neurons and noradrenergic innervation have been fully described for the rat brain ~9'29'33, the precise distribution of noradrenergic system in the human brain is not completely known. However, and in analogy with the rat brain, the human adrenaline neurotransmitter system arises from the C ~ and C 2 brainstem regions m mainly the locus ceruleUS 15'22'25'31'38-40'52. Using a sensitive phenylethanolamine N-methyltransferase (PNMT) assay, it has been demonstrated that in the human brain these neurons also project to forebrain regions, such as neocortex, hippocampus, amygdala, thalamus and hypothalamus, and to other hindbrain areas, such as several brainstem nuclei, cerebellum or spinal cord 7. In general, the anatomical distribution of a2-receptors as labelled with [3H]bromoxidine already described here follows the distribution of the presynaptic adrenergic terminals. In fact, areas that present the most important densities of a2-receptors, such as layer I of the neocortex, C A ~ stratum lacunosum-moleculare - - and dentate gyrus at the hippocampal formation, nuclei centralis, medialis and corticalis at the amygdaloid complex, anterior thalamic group and paraventricularis nuclei, several hypothalamic nuclei, substantia innominata, superior colliculus - - stlatum zonale --, lateral periaqueductal area, locus ceruleus, nucleus dorsalis of vagus, nucleus tractus solitarii, spinal cord dorsal horn - - substantia gelatinosa - - and stratum granularis in the cerebellar cortex, also show the highest density of brain adrenergic innervation 1'~'2'~.On the other hand, areas such as the globus pallidus or caudate-putamen-accumbens v~'2°, with low or negligible ['~H]bromoxidine specific bindiug, also present very scarce adrenergic innervation in the mammalian brain. The close correlation between the distribution and density of a2-receptors and the known distribution of adrenergic innervation of the human brain lends further weight to the proposed predominant presynaptic localization of a2-receptors 24. However, these data are not fully against a postsynaptic localization of a2-adrenoceptors, and carefully planned experiments utilizing for instance specific lesions are needed to adequately dissect the specific relationship of these receptors in specific regions of the CNS. The localization of a2-adrenoceptors in the human brain provides an anatomical basis for the known pharmacological effects of centrally acting a2-compounds. As extensively reviewed by Unnerstall et ai. for the rat brain 49, brainstem a2-receptors exert an important

126

modulatory influence on autonomic functions and integrate somatosensory and/or affective functions with autonomic mechanisms. For instance, a,-receptors at the medulla oblongata are involved in cardiovascular control mechanisms, justifying the known effects of clonidine as a central antihypertensive agent. On the other hand, pharmacological and biochemical studies point to a significant an6anxiety effect of clonidine in humans, an effect probably mediated by suppression of noradrenergie activity of the locus ceruleus neurons 44. Locus ceruleus projections to cortical forebrain areas are thought to be important in regulating some aspects of higher intellectual functions such as attention, memory, emotion, behaviour and of human motor function 2's'6. In Alzheimer's disease, an entity where a decrease in a2-adrenoceptors has been tentatively described in some forebrain areas using membrane binding methods ~2'46, the degeneration of locus ceruleus noradrenergic neurons has been related to the deficit in these higher brain functions 2'4'~''2°. Moreover, in different parkinsonisms, such as Parkinson's disease or supranuclear progressive palsy, movement disorders where the locus ceruleus degeneration is fairly constant j"'4s, reductions in the density of a2-receptors have been reported in the locus ceruleus projection areas in frontal cortex membrane homogenates z~ or in isolated cases by using autoradiographic methods and human tissue sections '~. The relevance of the here described anatomical distribution of a,-adrenoceptors in the human brain for a better knowledge of the neurochemistry of these human neurodegcnerative disorders needs not to be overemphasized. Arknowh,dl#nems, Our thanks to Dr. Jos6 Berclano for his excellent technical help in preparing this manuscript. This work was supported in part by a grant from CAYCIT, Ministry of Education (852/84),

REFERENCES I Ahlquist, R.P., A study of the adrenotropic receptors, Am. J, Physiol. 153 (1948) 586-f~00, 2 Agid, Y., Javoy-Agid, F. and Ruberg, M., Biochemistry of neurotransmitters in Parkinson's disease, in C.D. Marsden and S, Fahn (Eds.), Morement disorders 2, 1987, Butterworths, London, pp, 166-231. 3 Asakura, M,, Tsukamoto, T., lmafuku, J,, Matsui, H,, Ino, M, and Hasegawa, K., Quantitative analysis of rat brain c~rreceptors discriminated by [~H]clonidine and [~H]rauwolscine, E,r, J, Pharmacol., 106 (1985) 141 - 147. 4 Bondareff, W., Mountjoy, C,Q, and Roth, M,, Loss of adrcnergic neurons of origin of the adrenergic projection to cerebral cortex (nucleus locus ceruleus) in senile dementia, Neurolo~, 32 (1982) 164-168. 5 Burke, W.J,, Chang, H,D,, Huang, J,S,, Huang, S,H,, flaring, J.H., Marshall, G.L, and Job, T,H,, Evidence for retrograde degeneration of epinephrine neurtms in AIzheimer's disease, Am~, Neurol., 24 (1988) 532-536. 6 Burke, W.J., Chung, H,D, and Nakra, R,R,S,, Phenylethanol-

amine N-methyltransferase is decreased in Alzheimer's disease brains, Ann. Neuroi., 22 (1987) 278-280. 7 Burke, W.,[., Hanson, D.M. and Chung, H.D., A highly sensitive assay for phenylethanolamine N-methyltranferase in human brain, Proc. Soc. Exp. Biol. Med., 181 (1986)66-70. 8 Cambridge, D., UK-14304, a potent and selective a2-agonist for the characterization of a-adrenoceptor subtypes, Eur. J. Pharmacol., 72 (1981) 413-415. 9 Carlson, M.J. and Andorn, A.C. [3H]Clonidine binds at multiple high affinity states in human prefrontal cortex, Eur. J. Pharmacol., 123 (1986) 73-78. l0 Cash, R., Raisman, R., Ruberg, M. and Agid, Y., Adrenergic receptors in frontal cortex in human brain, Eur. J. Pharmacoi., 108 (1985) 225-232. 11 Cash, R., Ruberg, M., Raisman, R. and Agid, Y., Adrenergie receptors in Parkinson's disease, Brain Res., 322 (1984) 269-275. 12 Cross, A.J., Crow. T.J'., Johnson, J.A., Perry, E.K., Perry, R.P., Blessed, G. and Tomlinson, E., Studies on neurotransmitter receptor systems in neocortex and hippocampus in senile dementia of the AIzheimer type. J. Neurol. Sci., 64 (1984) 109-117, 13 Dausse, J.P., Quan-Bui, K.H.L. and Meyer, P. a I and a2-adrenoceptors in rat cerebral cortex: effects of neonatal treatment with 6-hydro~dopamine, Fur. J, Pharmacol., 78 (1982) 15-20. 14 DeLangen, C.D.J., Hogenboom, F. and Mulder, A.H., Presynaptic noradrenergic a-receptors and modulation of 3H-noradrenaline release from rat brain synaptosomes, Fur. J. Pharmacol., 60 (1979) 79-89. 15 Fallon, J.H.. Koziell, D.A. and Moore, R.Y. Catecholamine innervation of the basal forebrain: amygdala, suprarhinai and entorhinal cortex, J. Comp. Neurol., 180 (1978) 533-544. 16 Forno. L.S., Pathology of Parkinson's disease, in C,D. Marsden and S. Fahn (Eds.), Mot,ement disorders 2, Butterworths, London, 1987, pp. 25-30. 17 Harrison, J.K., Pearson, W.R. and Lynch, K.R., Molecular characterization of al and a:-adrenoceptors, Trends Pharmnacol. Sci., 12 (1991) 62-67. 18 Itaeusler, G., CIonidine induced inhibition of sympathetic nerve activity: no indication for a central presynapti¢ or indirect sympathomimetic mode of action, Naunyn Schmiedeberg~' Arch, Phar. macol,, 286 (1974) 9%,111, 19 HOkfelt, T., Johansson, O, and Ooldsteln, M,, Central cute. cholamine neurons as revealed by immunohistochemistry with special reference to adrenaline neurons, in A, Bj(irklund and T, lt6kfelt (Eds,), tlandhc~k of Chemical Neueoanatomy, Vol, 2, Cla,'sicai Transmitters in Itle CNS, Elsevier, Amsterdam, 1984, pp. 157-276. 20 Iversen, L,L,, Rossor, M,N,, Reynolds, G,P,, Hills, R,, Roth, M,, Mounqoy, C,Q,, Foote, $,L,, Morrison, J,H, and Bloom, F,E,, Loss of pigmented dopamine beta-hydro~lase positive cells from locus ceruleus in senile dementia of AIzheimer type, Neuroari, Lett., 39 (1983) 95-100. 21 Kobinger, W, and Pichler, W,, The central modulatory effect of clonidine on the cardiopressor reflex after suppression of synthesis and storage of noradrenaline, £ur, 7, Phamm¢ol,, 30 (1975) 31 !-320, 22 Kopp, N,, Denoroy, L,, Renaud, B,, Pujol, J,F., Tabib, A, and Tommasi, M, Distribution of adrenaline-synthesizing enzyme activity in the human brain, 7, Neurol, Sci., 41 (1979) 397-409. 23 Kuhar, M,J, and Unnerstall, J,R,, Quantitative receptor mapping by autoradiography: some current technical problems. Trends Neueos('i,, 8 (1985) 49, 24 Langer, S,Z,, Presynaptic regulation of the release of catecholamines, Pharmacol, Re~',, 32 (1980) 337-362, 25 Mann, D,M,A,, Yates, P,O, and Hawkes, J,, Pathology of the human locus ceruleus, Clin, Neuropathol,, 2 (1983) 1-7. 26 Meana, J,J, and Garch-Sevilla, J,A,, Increased az-adrenoceptor density in the frontal cortex of depressed suicide victims, J, Neural Transm,, 70 (1987)377-381. 27 Meana, J,J,, Barturen, F, and Garc[a Sevilla, J,A,, Characterization and regional distribution of a2-adrenoceptors in postmortem human brain using the full agonist 3H-UK-14304, J, Neurochem., 52 (1989) 1210-1218,

127 28 Minneman, K.P., Piffman, R.N. and Molinoff, P.B., Adrenergic receptor subtypes: Properties, distribution and regulation, Annu. Rev. Neurosci., 4 (1981) 419-462. 29 Moore, R.Y.and Card, J.P., Noradrenaline-containing neuron systems. In A. Bj6rklund and T. H6kfelt (Eds.), Handbook of Chemical Neuroanatomy, Vol. 2, Classical Transmitters in the CNS, Elsevier, Amsterdam, 1984, pp. 123-156. 30 Morris, MJ., Elghozi, J.L., Dausse, J.L. and Meyer, P., al and a 2 adrenoceptors in rat cerebral cortex: effect of frontal Iobotomy. Naunyn-Schmiedeberg's Arch. Pharmacol., 316 (1981) 42-44. 31 Olson, L., Nystr6m, B. and Seiger, A., Monoamine fluorescence histochemistry of human postmortem brain, Brain Res., 63 (1973) 317-320. 32 Palacios, J.M., Hoyer, D. and Cort6s, R., a~-Adrenoceptors in the mammalian brain: similar pharmacology but different distribution in rodents and primates, Brain Res., 419 (1987) 65-75. 33 Parent, A., Poitras, D. and Dub(~, L., Comparative anatomy of central monoaminergic systems. In A. Bj6rklund and T. H6kfelt (Eds.), Handbook of Chemical Neuroanatomy, Vol. 2, Classical Transmitters in the CNS, Elsevier, Amsterdam, 1984, pp. ~09-440. 34 Pascual, J., Gonz;ilez, A,M., Berciano, J.A., Sedano, M.I., Figols, J., Fl6rez, J. and Pazos, A., General decrease of a,-adrenoceptots in the brain of a patient with progressive supranuclear palsy: an autoradiographic study, J. Neurol., 235 (suppl.) (1988) 129. 35 Pascual, J., del Arco, C., Gonzfilez, A.M., Diaz, A., del Olmo, E. and Pazos, A., Regionally specific age-dependent decline in a2adrenoceptors: an autoradiographic study in human brain, Neurosci Lett., 133 (1991) 279-283. 36 Pazos, A., Gonzfilez, A.M., Pascual, J., Meana, J.J., Barturen, F. and Garc|a-Sevilla, J,A., a2-Adrenoceptors in human forebrain: autoradiographic visualization and biochemical parameters using the agonist [3H]UK-14304, Brain Res., 475 (1988) 361-365. 37 Pazos, A. and Palacios, J.M,, Neurotransmitter receptor mapping in human post mortem brain by autoradiography, in M. Lewis and N.A. Shariff (Eds.), Brain Imaging: Techniques and Applica. tions, Horwood, Chichester, 1989, pp. 110-128. 38 Pearson, J., Goldstein, M. and Brandeis, L, Tyrosine hydrox'ylase immunohistochemistry in human brain, Brain Res., 165 (1979) 333-337. 39 Pearson, J,, Goldste.in, M., Markey, K. and Brandeis, L, Huma,! brainstem catecholamine neuronal anatomy as indicated by immunocytoeh©mistry with antibodies to tyrosine hydro~lase, Neuroscience, 8 (I 983) 3-32. 40 Plekel, V.M., Segal, M. and Bloom, F.E., A radioautoradiographic study of the efferent pathways of the nucleus locus eeruleus, J. Comp. Neurol., 155 (1974) 15-42. 41 Pelayo, F., Dubovich, M.L. and Langer, S.Z., Inhibition of neuronal uptake reduces the presynaptic effect of donidine but not of a-methylnt~radrenaline on the stimulation-evoked release of

42

43

44 45 46

47 48 49

50

51

52 53 54

[3H]noradrenaline from rat occipital cortex slices, Eur. Jr. Pharmacol., 64 (1980) 143-155. Probst, A., Cortes, R. and Palacios, J.M., Distribution of a,adrenergic receptors in the human brainstem: an autoradiographic study using 3H-p-aminoclonidine, Fur. J. Pharmacol., 106 (1984) 477-488. Rainbow, T.C., Biegon, A. and Berck, D.J., Quantitative autoradiography with tritium labelled ligands: comparison of biochenfical and densitometric measurements, J. Neurosci. Methods, 11 (1984) 231-241. Redmon, D.E., Does clonidine alter anxiety in humans?, Trends Pharmacol. Sci., 3 (1982) 477-480. Ruffolo, R.R., Stereochemical requirements for activation and blockade of oq and a2-adrenoceptors, Trend Pharmacol. Sci., 5 (1984) 160-164. Simohama, S., Taniguchi, T., Fujiwara, M. and Kameyama, M., Biochemical characterization of a-adrenergic receptors in human brain and changes in Alzheimer-type dementia, J. Neurochem., 47 (1986) 1295-1301. Starke, K., Regulation of noradrenaline release by presynaptic receptor systems, Ret,. Physiol. Biochem. Pharmacol., 77 (1977) 1-124. Steele, J.C. Progressive supranuclear palsy. In PJ. Vinken and G.W. Bruyn (Eds.), Handbook of Clinical Neurology, Vol. 22, North Holland, Amsterdam, 1975, pp. 217-229. Unnerstall, J.R., Kopajtic, T.A. and Kuhar, M.J., Distribution of alpha2 agonist binding sites in the rat and human central nervous system: analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents, Brain Res. Rev., 7 (1984) 69-101. Unnerstall, J.R., Niehoff, D.L., Kuhar, MJ. and Palacios, J.M., Quantitative receptor autoradiography using 3H-ultrofilm: applications to multiple benzodiazepine receptors, J. Neurosci. Methods, 6 (1982) 59-73. U'Prichard, D.C., Reisine, T.D., Mason, S.T., Fibiger, H.C. and Yamamura, H.I., Modulation of rat brain a- and ~-adrenergic receptor populations by lesion of the dorsal noradrenergic bundle, Brain Res., 187 (1980) 143-154. Waterhouse, B.D. Chia-Sheng, L., Burne, R.A, and Woodward, D.J., The distribution of neocortical projection neurons in the locus ceruleus, J. Comp. Neurol., 217 (1983) 418-431. Weinreich, P. and Seeman, P., Binding of adrenurgic ligands (['~l*i]clonidine and ['~ioI[WB-4101) to multiple sites in human brain, Biochem. Phannaeol., 30 (1981) 3115-3120. Wemer, J., Frankhuzzen, A.L. and Mulder. A.H., Pharmacological characterization of presynaptic a.adrcnoceptors in the nucleus tractus solitarii and the cerebral cortex of the rat, Neuropharmacology 21 (1982) 499-506.