0306-4522/91 $3.00 + 0.00
Neuroscience Vof. 44, No. 2, pp. 4%.5-481,1991
Pergainon Press pk 0 1991IBRO
Printedin Great Britain
HISTAMINE NEURONS IN HUMAN HYPOTHALAMUS: ANATOMY IN NORMAL AND ALZHEIMER DISEASED BRAINS M. S. AIRAKSINEN,* A. PAISTAU,~L. PAW~RVI,~ K. ~I~I~~~N,§ R. SUOMALA~NEN\\ and
P. RIEKKINEN,$
P. PANULA*~
Departments of *Anatomy and tPathology. University of Helsinki, Sil~vuo~n~nger %A, SF-001 70 Hetsinki, Finland Departments of ~Pathology and $Neurology, Kuopio University Central Hospital, 70210 Kuopio, Finland j/Department of pathology, Kiveli Hospital, 00260 Heisinki, Finland Ahatmct-The anatomy of histamin~immunor~ct~ve cell bodies in normal adult human brain was examined in detail. In addition, the distribution of these cells in three cases of Alzheimer’s disease was compared to the distribution of neurofibrillary tangles. Histamine-immunoreactive cell bodies were confined to the tuberal and posterior hypothalamus, forming the tuberomammillary nuclear complex. Most of the about 64,000 histamine neurons were large and multipolar. They comprised four distinct parts: (i) a major ventral part corresponding to the classical tuberomammillary nucleus, (ii) a medial part including the supramammillary nucleus and part of the posterior hypothalamic area, (iii) a caudal ~r~ammilla~ part, and {iv) a minor lateral part. The parts showed some similarity wifh the subgroups in rat. In human, as compared to rat, the histamine neurons occupy a larger proportion of the hypothalamus. Numerous neurofib~ll~y tangles were found in the Alzheimer hypothalami, ~oncentra~d in the tu~romammil~~ area. Most of them were of globular type and extracellular, and only a mino~ty were histamine immunor~ctive. They may represent remnants of degenerated tu~romammilla~ neurons.
Histamine
is a diamine
neurotransmitter
involved
in
many basic functions (e.g. Ref. 102 for a review). In all adult mammals studied, histamine-containing neurons are located in the tuberomammillary (TM) neuron complex of the posterior hypothalamus and have fiber projections to most brain areas.” In contrast, the developing rat7 and adult lamprey9 have histamine-immunoreactive (-IR) neurons also in the medulla. Recently, we reported on a histaminergic system in the human brain with nerve fibers in the cerebral cortex and cell bodies in the hypothalamus.~ Ifowever, detailed distribution of the cell bodies or their possible presence in other brain areas was not studied. The histamine neurons were first localized in the rat posterior hypothalamus in three magnocellular nuclei of Bleier et al.,’ namely the postmammillary caudal (PMC), caudal (CMC) and tuberal (TMC) magnocellular nuclei.56*80*M Because the neurons of these three nuclei and some neurons scattered in adjacent hypothalamus are mo~hoio~cally similar --...
..-_
TTo whom correspondence should be addressed. A~~r~j~t~o~: Ag, Alzheimer’s disease; CMC, caudat magnocelh&u nucleus: DAB. di~ino~~dine~ EDCDI. l~ethyl-3(3-dimet~ylami~opro~~l)~rb~iim~de; -IR; -immunoreactive; MAO, monoamine oxidase; NFT, neurofibrillary tangle(s); PBS, phosphate-buffered saline; PMC, postmammillary caudal nucleus; REM, rapid eye movement; THA, 1,2,3,4-tetrahydroamino-9-acridine; TM, tuberomammillary (nucleus); TMC, tuberal magnocelhdar nucleus.
and seemed to form one continuous cell group, as originally suggested by Diepen,‘* Kiihler et a1.37referred to them as the TM nucleus. The nucleus was divided into the ventral (corresponding to PMC and CMC), dorsomedial (corresponding to TMC) and diffuse parts. Recently, the morphology of the TM neuron complex in the rat brain has been extensively analysed by using different immunohistochemi~i markers and anatomically divided into five sub~ou~s.‘~,~.~’ Ericson et ~1.” divided it into ventral, medial and diffuse parts, and further the ventral part into caudat and rostral components, and the medial part into ventral and dorsal components. However, tract-tracing studies in the rat have indicated that efferent projections from’q.30,37*81,86 and afferent projections to20.g7U9” different parts of the TM neuron complex are similar between the subgroups, i.e. no functional difference between them seems to exist. In this study, we examined the normal anatomy of histamine-IR nerve cell bodies in the human brain, and mapped their exact locations in the hypothalamus. We also surveyed for their possible presence in other areas of the brainstem. The pathology of choline& neurons in the nucleus basalis of Meynert, serotonergic neurons in the raphe nuclei and noradrenergic neurons in locus coeruleus in patients with Alzheimer’s disease (AD) is established (e.g. Ref. 27 for a review). The presence of neurofibrillary tangles (NFT) in the posterior hypothalamus of AD patients,2q*31especially in the large neurons of the TM area,72.76suggested that the
466
M. S.
AIRAKSMENel ul.
histaminergic system might also be involved in this disease. However, reports on histamine levels in AD have been conflicting. i1.47Therefore, we studied the distribution of histamine-IR neurons in three welfdocumented cases of AD, and compared it to that of age-matched controls and to the distribution of NFT in the same sections. A preliminary report of this work has been presented.4 EXPERIMENTAL
PROCEDURES
Cell counting
To obtain a crude estimate of the mean number of histamine@ cell bodies in normal adult human brain, the formula of Abercrombie’ was used to correct counted split cell bodies. Briefly, every 30th section was counted through the TM complex (at l-mm intervals, total eight sections/ case). The actual number of histamine-IR cell bodies (N,) in each case (Nl-3) was calculated from the equation (N, = 3O*N, x r/(t + D), where N, is the total number of histamine-IR neuronal profiles .counted, t = 35 f 5 pm (section thickness) and D = 27 & 8 pm (mean + S.E.M. diameter of the ceil bodies).
Post mortem sampks
Brain atlas
Brain tissues were obtained from routine autopsies of three AD (All3) and six non-neurological (age-matched controls = Cll3, normal adults = Nl-3) patients. Table 1 shows the clinical and neuropathological data of the cases. Only the right side of each AD hypothalamus was fixed for this study. The whole brainstem of one normal case (N2) was examined for histamine-IR cell bodies.
Several human brain atlases were compared to locate the hypothalamic nuclei.‘2J8~74~70 The abbreviations were adopted from the rat atlas of Paxinos and Watson.“’ The maps were drawn using a camera lucida from the brain sections used in this study.
Tissue fixation and staining
The tissues were cut into about l-cm-thick slices and fixed by immersion in 4% I-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCDI; Sigma, U.S.A.) in 0.1 M phosphate buffer (PH 7.0) for two to four days at 4°C. After one to two days, the EDCDI-fixative was changed. Then the tissues were placed into 20% sucros~phosphate-bu~ered saline (PBS) solution and kept for three to five days at 4°C until they sank. Frontal, sagittal or horizontal sections (3&40 pm thick) were cut with a cryostat on gelatin-coated glass slides and stored in tightly closed boxes at -20°C. Before immunohistochemistry, the sections on the slides were fixed for 20min in 4% EDCDI. The histamine-antiserum was used at dilution 1:200&l : 5000.Specificity of the antiserum and details of the staining procedure including the peroxidase-antiperoxidase method were described earlier.@ All staining was blocked by preabsorbing the diluted antiserum with histamine-ovalbumin conjugates (1 pg/ml) but not with L-histidine-ovalbumin (< 100 p&/ml) or other conjugates (data not shown). Some sections were counterstained with Luxol Fast Blue (Sinma, U.S.A.) and Cresyl Violet to locate myelinated and-nuclear structures. The method of Gordonz5 was applied to enhance contrast in diaminobenzidine (DAB) photomicrography. To locate the NFT, some sections were fixed with 10% paraformaldehyde for five days, washed overnight with PBS at 4”C, stained for 10min in 1% thiogavin-S (Sigma, U.S.A.) solution,” differentiated in 80% alcohol, washed twice in water. and mounted with glycerol-PBS solution. Some of these sections were also treated with hematoxylin (Merck, U.S.A.) to stain the nuclei. Table I. Clinical and neuropathological Case Nl N2 N3 Cl C2 ::$ A2 A3
Age
Sex
48 49 49 75 90 86 85 87 78
F F M M F F F F F
Cause of death Breast cancer Breast cancer Myocardial infarction Myocardial infarction Myocardial infarction Colon cancer Dementia for 12 years Dementia for 15 years Dementia for three years
Effect of different post mortem intervals, i.e. period between death and start of fixation, on the detection of histamine cell bodies was tested by means of an animal experiment. Two female rats (R1,2, weight 17@180 g) were killed by pentobarbital overdose (1 mg/lOO g, i.p.) and kept at room temperature for 2 h and a further 22 h (Rlf or 46 h (R2) at 4°C. Then the brains were removed, fixed by immersion for 24 h in EDCDI at 4°C and nrocessed for histamine immunohistochemistry. The number of histamine-IR cell bodies in these samples was compared to those obtained from perfusion-fixed brains.58 RESULTS
distributive
qf
h~ta~i~e
neurons
in normal
hM~an
brain
Regardless of surveying the whole brainstem (of case N2) no histamine-IR cell bodies were found except in the posterior hypothalamus. The location of the cell bodies in a normal adult human hypothalamus is shown in Figs l-3 in frontal (case Ni), horizontal (right side of case N3) and sagittal sections (left side of case N3). Analysis of the sections showed that the histamine-IR neurons formed a continuous cell system, designated as the TM neuron complex. It extended from the supraoptic nucleus and the preoptic area rostrally to the ventral tegmental area and the substan~a nigra caudally, from the cerebral peduncle and the internal capsule laterally to the
data of patients with Alzheimer-type dementia and controls PM interval* (h) 70 46 48 40 48 12 40 20 39
HC ND ND O/O + +/+ 0/+++ ojo ++/+++ +i-+/t++ ++/+i+
Plaques/tangiest TC ND ND O/O O/O ND O/O ++/+ +++I+++ ++t/+++
HT
010 O/O O/O -t/O 0/+ o/o +i+++ o/+ + + ++/+++
*Period between death and autopsy, all cadavers stored at 4°C. TO, + , + + , + + + = O1one to seven, eight to 19, > 20 plaques or tangles per microscope (Leitz Aristoplan) field in x 40 or x 100 magnificatrons, respectively. $With multiple infarcts. HC, hippocampus (CAL); TC, temporal cortex (medial gyrus); HT. hypothalamus (posterior part); C, N, controls; A, Alzheimer-TV dementia; ND, not done.
467
Histamine neurons in human hypothalamus
P
L
GP
LH
LH
I I *-
:*
VMH
..
. .. :i
I
*. . ‘X.
. LTu
c-
Neuroscience Vof. 44, No. 2, pp. 4%.5-481,1991
Pergainon Press pk 0 1991IBRO
Printedin Great Britain
HISTAMINE NEURONS IN HUMAN HYPOTHALAMUS: ANATOMY IN NORMAL AND ALZHEIMER DISEASED BRAINS M. S. AIRAKSINEN,* A. PAISTAU,~L. PAW~RVI,~ K. ~I~I~~~N,§ R. SUOMALA~NEN\\ and
P. RIEKKINEN,$
P. PANULA*~
Departments of *Anatomy and tPathology. University of Helsinki, Sil~vuo~n~nger %A, SF-001 70 Hetsinki, Finland Departments of ~Pathology and $Neurology, Kuopio University Central Hospital, 70210 Kuopio, Finland j/Department of pathology, Kiveli Hospital, 00260 Heisinki, Finland Ahatmct-The anatomy of histamin~immunor~ct~ve cell bodies in normal adult human brain was examined in detail. In addition, the distribution of these cells in three cases of Alzheimer’s disease was compared to the distribution of neurofibrillary tangles. Histamine-immunoreactive cell bodies were confined to the tuberal and posterior hypothalamus, forming the tuberomammillary nuclear complex. Most of the about 64,000 histamine neurons were large and multipolar. They comprised four distinct parts: (i) a major ventral part corresponding to the classical tuberomammillary nucleus, (ii) a medial part including the supramammillary nucleus and part of the posterior hypothalamic area, (iii) a caudal ~r~ammilla~ part, and {iv) a minor lateral part. The parts showed some similarity wifh the subgroups in rat. In human, as compared to rat, the histamine neurons occupy a larger proportion of the hypothalamus. Numerous neurofib~ll~y tangles were found in the Alzheimer hypothalami, ~oncentra~d in the tu~romammil~~ area. Most of them were of globular type and extracellular, and only a mino~ty were histamine immunor~ctive. They may represent remnants of degenerated tu~romammilla~ neurons.
Histamine
is a diamine
neurotransmitter
involved
in
many basic functions (e.g. Ref. 102 for a review). In all adult mammals studied, histamine-containing neurons are located in the tuberomammillary (TM) neuron complex of the posterior hypothalamus and have fiber projections to most brain areas.” In contrast, the developing rat7 and adult lamprey9 have histamine-immunoreactive (-IR) neurons also in the medulla. Recently, we reported on a histaminergic system in the human brain with nerve fibers in the cerebral cortex and cell bodies in the hypothalamus.~ Ifowever, detailed distribution of the cell bodies or their possible presence in other brain areas was not studied. The histamine neurons were first localized in the rat posterior hypothalamus in three magnocellular nuclei of Bleier et al.,’ namely the postmammillary caudal (PMC), caudal (CMC) and tuberal (TMC) magnocellular nuclei.56*80*M Because the neurons of these three nuclei and some neurons scattered in adjacent hypothalamus are mo~hoio~cally similar --...
..-_
TTo whom correspondence should be addressed. A~~r~j~t~o~: Ag, Alzheimer’s disease; CMC, caudat magnocelh&u nucleus: DAB. di~ino~~dine~ EDCDI. l~ethyl-3(3-dimet~ylami~opro~~l)~rb~iim~de; -IR; -immunoreactive; MAO, monoamine oxidase; NFT, neurofibrillary tangle(s); PBS, phosphate-buffered saline; PMC, postmammillary caudal nucleus; REM, rapid eye movement; THA, 1,2,3,4-tetrahydroamino-9-acridine; TM, tuberomammillary (nucleus); TMC, tuberal magnocelhdar nucleus.
and seemed to form one continuous cell group, as originally suggested by Diepen,‘* Kiihler et a1.37referred to them as the TM nucleus. The nucleus was divided into the ventral (corresponding to PMC and CMC), dorsomedial (corresponding to TMC) and diffuse parts. Recently, the morphology of the TM neuron complex in the rat brain has been extensively analysed by using different immunohistochemi~i markers and anatomically divided into five sub~ou~s.‘~,~.~’ Ericson et ~1.” divided it into ventral, medial and diffuse parts, and further the ventral part into caudat and rostral components, and the medial part into ventral and dorsal components. However, tract-tracing studies in the rat have indicated that efferent projections from’q.30,37*81,86 and afferent projections to20.g7U9” different parts of the TM neuron complex are similar between the subgroups, i.e. no functional difference between them seems to exist. In this study, we examined the normal anatomy of histamine-IR nerve cell bodies in the human brain, and mapped their exact locations in the hypothalamus. We also surveyed for their possible presence in other areas of the brainstem. The pathology of choline& neurons in the nucleus basalis of Meynert, serotonergic neurons in the raphe nuclei and noradrenergic neurons in locus coeruleus in patients with Alzheimer’s disease (AD) is established (e.g. Ref. 27 for a review). The presence of neurofibrillary tangles (NFT) in the posterior hypothalamus of AD patients,2q*31especially in the large neurons of the TM area,72.76suggested that the
466
M. S.
AIRAKSMENel ul.
histaminergic system might also be involved in this disease. However, reports on histamine levels in AD have been conflicting. i1.47Therefore, we studied the distribution of histamine-IR neurons in three welfdocumented cases of AD, and compared it to that of age-matched controls and to the distribution of NFT in the same sections. A preliminary report of this work has been presented.4 EXPERIMENTAL
PROCEDURES
Cell counting
To obtain a crude estimate of the mean number of histamine@ cell bodies in normal adult human brain, the formula of Abercrombie’ was used to correct counted split cell bodies. Briefly, every 30th section was counted through the TM complex (at l-mm intervals, total eight sections/ case). The actual number of histamine-IR cell bodies (N,) in each case (Nl-3) was calculated from the equation (N, = 3O*N, x r/(t + D), where N, is the total number of histamine-IR neuronal profiles .counted, t = 35 f 5 pm (section thickness) and D = 27 & 8 pm (mean + S.E.M. diameter of the ceil bodies).
Post mortem sampks
Brain atlas
Brain tissues were obtained from routine autopsies of three AD (All3) and six non-neurological (age-matched controls = Cll3, normal adults = Nl-3) patients. Table 1 shows the clinical and neuropathological data of the cases. Only the right side of each AD hypothalamus was fixed for this study. The whole brainstem of one normal case (N2) was examined for histamine-IR cell bodies.
Several human brain atlases were compared to locate the hypothalamic nuclei.‘2J8~74~70 The abbreviations were adopted from the rat atlas of Paxinos and Watson.“’ The maps were drawn using a camera lucida from the brain sections used in this study.
Tissue fixation and staining
The tissues were cut into about l-cm-thick slices and fixed by immersion in 4% I-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCDI; Sigma, U.S.A.) in 0.1 M phosphate buffer (PH 7.0) for two to four days at 4°C. After one to two days, the EDCDI-fixative was changed. Then the tissues were placed into 20% sucros~phosphate-bu~ered saline (PBS) solution and kept for three to five days at 4°C until they sank. Frontal, sagittal or horizontal sections (3&40 pm thick) were cut with a cryostat on gelatin-coated glass slides and stored in tightly closed boxes at -20°C. Before immunohistochemistry, the sections on the slides were fixed for 20min in 4% EDCDI. The histamine-antiserum was used at dilution 1:200&l : 5000.Specificity of the antiserum and details of the staining procedure including the peroxidase-antiperoxidase method were described earlier.@ All staining was blocked by preabsorbing the diluted antiserum with histamine-ovalbumin conjugates (1 pg/ml) but not with L-histidine-ovalbumin (< 100 p&/ml) or other conjugates (data not shown). Some sections were counterstained with Luxol Fast Blue (Sinma, U.S.A.) and Cresyl Violet to locate myelinated and-nuclear structures. The method of Gordonz5 was applied to enhance contrast in diaminobenzidine (DAB) photomicrography. To locate the NFT, some sections were fixed with 10% paraformaldehyde for five days, washed overnight with PBS at 4”C, stained for 10min in 1% thiogavin-S (Sigma, U.S.A.) solution,” differentiated in 80% alcohol, washed twice in water. and mounted with glycerol-PBS solution. Some of these sections were also treated with hematoxylin (Merck, U.S.A.) to stain the nuclei. Table I. Clinical and neuropathological Case Nl N2 N3 Cl C2 ::$ A2 A3
Age
Sex
48 49 49 75 90 86 85 87 78
F F M M F F F F F
Cause of death Breast cancer Breast cancer Myocardial infarction Myocardial infarction Myocardial infarction Colon cancer Dementia for 12 years Dementia for 15 years Dementia for three years
Effect of different post mortem intervals, i.e. period between death and start of fixation, on the detection of histamine cell bodies was tested by means of an animal experiment. Two female rats (R1,2, weight 17@180 g) were killed by pentobarbital overdose (1 mg/lOO g, i.p.) and kept at room temperature for 2 h and a further 22 h (Rlf or 46 h (R2) at 4°C. Then the brains were removed, fixed by immersion for 24 h in EDCDI at 4°C and nrocessed for histamine immunohistochemistry. The number of histamine-IR cell bodies in these samples was compared to those obtained from perfusion-fixed brains.58 RESULTS
distributive
qf
h~ta~i~e
neurons
in normal
hM~an
brain
Regardless of surveying the whole brainstem (of case N2) no histamine-IR cell bodies were found except in the posterior hypothalamus. The location of the cell bodies in a normal adult human hypothalamus is shown in Figs l-3 in frontal (case Ni), horizontal (right side of case N3) and sagittal sections (left side of case N3). Analysis of the sections showed that the histamine-IR neurons formed a continuous cell system, designated as the TM neuron complex. It extended from the supraoptic nucleus and the preoptic area rostrally to the ventral tegmental area and the substan~a nigra caudally, from the cerebral peduncle and the internal capsule laterally to the
data of patients with Alzheimer-type dementia and controls PM interval* (h) 70 46 48 40 48 12 40 20 39
HC ND ND O/O + +/+ 0/+++ ojo ++/+++ +i-+/t++ ++/+i+
Plaques/tangiest TC ND ND O/O O/O ND O/O ++/+ +++I+++ ++t/+++
HT
010 O/O O/O -t/O 0/+ o/o +i+++ o/+ + + ++/+++
*Period between death and autopsy, all cadavers stored at 4°C. TO, + , + + , + + + = O1one to seven, eight to 19, > 20 plaques or tangles per microscope (Leitz Aristoplan) field in x 40 or x 100 magnificatrons, respectively. $With multiple infarcts. HC, hippocampus (CAL); TC, temporal cortex (medial gyrus); HT. hypothalamus (posterior part); C, N, controls; A, Alzheimer-TV dementia; ND, not done.
467
Histamine neurons in human hypothalamus
P
L
GP
LH
LH
I I *-
:*
VMH
..
. .. :i
I
*. . ‘X.
. LTu
c-
