The von Economo neurons (VENs) are large bipolar neurons located in frontoinsular (FI) and anterior cingulate cortex in great apes and humans, but not other primates. We performed stereological counts of the VENs in FI and LA (limbic anterior, a component of anterior cingulate cortex) in great apes and in humans. The VENs are more numerous in humans than in apes, although one gorilla approached the lower end of the human range. We also examined the ontological development of the VENs in FI and LA in humans. The VENs first appear in small numbers in the 36th week post-conception, are rare at birth, and increase in number during the first 8 months after birth. There are significantly more VENs in the right hemisphere than in the left in FI and LA in postnatal brains of apes and humans. This asymmetry in VEN numbers may be related to asymmetries in the autonomic nervous system. The activity of the inferior anterior insula, which contains FI, is related to physiological changes in the body, decisionmaking, error recognition, and awareness. The VENs appear to be projection neurons, although their targets are unknown. We made a preliminary study of the connections of FI cortex based on diffusion tensor imaging in the brain of a gorilla. The VEN-containing regions connect to the frontal pole as well as to other parts of frontal and insular cortex, the septum, and the amygdala. It is likely that the VENs in FI are projecting to some or all of these structures and relaying information related to autonomic control, decision-making, or awareness. The VENs selectively express the bombesin peptides neuromedin B (NMB) and gastrin releasing peptide (GRP) which are also expressed in another population of closely related neurons, the fork cells. NMB and GRP signal satiety. The genes for NMB and GRP are expressed selectively in small populations of neurons in the insular cortex in mice. These populations may be related to the VEN and fork cells and may be involved in the regulation of appetite. The loss of these cells may be related to the loss of satiety signaling in patients with frontotemporal dementia who have damage to FI. The VENs and fork cells may be morphological specializations of an ancient population of neurons involved in the control of appetite present in the insular cortex in all mammals. We found that the protein encoded by the gene DISC1 (disrupted in schizophrenia) is preferentially expressed by the VENs. DISC1 has undergone rapid evolutionary change in the line leading to humans, and since it suppresses dendritic branching it may be involved in the distinctive VEN morphology.

Computer visualization techniques were used to map and to quantitatively reconstruct the entire locus coeruleus, including the nucleus subcoeruleus, to compare the topographic patterns of cell loss in postmortem brains from patients with Parkinson's disease, Alzheimer's disease, and Down syndrome. There was comparable cell loss in all three diseases (approximately 60%) compared with aged normal subjects, and there was a significant loss of nucleus subcoeruleus cells specifically in patients with Parkinson's disease (63%). There was a significant positive correlation between the magnitude of locus coeruleus cell loss and the duration of Alzheimer's disease, but no such correlation was found for Parkinson's disease. In patients with Parkinson's disease, there was comparable cell loss throughout the rostral-caudal extent of the nucleus; however, in patients with Alzheimer's disease and Down syndrome, the greatest cell loss always occurred within the rostral portion of the nucleus, with a relative sparing of caudal cells. These data are consistent with the hypothesis that cell loss in Parkinson's disease is the result of a pathological process that attacks the catecholaminergic cells of the locus coeruleus and the subcoeruleus in general; in Alzheimer's disease and Down syndrome, however, the pathological process only affects the rostral, cortical-projecting locus coeruleus cells and spares the caudal, noncortical-projecting cells.

Computer imaging and immunohistochemical staining techniques were used to determine which midbrain dopaminergic (DA) cells are spared in Parkinson's disease (PD), and in animals treated with the DA neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), and whether the spared cells contain the calcium-binding protein, calbindin-D28k (CaBP). The PD patients had more than 55% fewer midbrain DA neurons than age-matched normal subjects. The cell loss occurred within the combined substantia nigra and retrorubral area (greater than 61%; DA nuclei A9 and A8, respectively), and the ventral tegmental area (greater than 42%; DA nucleus A10). The cell loss was greatest within the ventral portion of the nucleus A9. A similar pattern of DA cell loss was observed in MPTP-treated Macaca fascicularis monkeys. The CaBP-containing cells were located specifically in the cell regions spared by PD and by MPTP-treatment in both monkeys and C57BL/6 mice. These data suggest that PD and MPTP both destroy the same population of midbrain DA neurons within nuclei A8, A9, and A10, and that perhaps CaBP protects the DA neurons from cell death caused by both PD and MPTP.

Quantitative neuroanatomical techniques were used to determine whether with aging there is random or systematic loss of locus coeruleus (LC) neurons in the human brain. The cells were identified by immunohistochemical staining for the catecholaminergic enzyme tyrosine hydroxylase and/or by neuromelanin pigment content. Cell locations were mapped, using computer imaging procedures, in horizontal sections spaced 0.5 to 0.8 mm throughout the rostrocaudal extent of the nucleus in 17 cases, from 1 to 104 years of age. Neuromelanin pigment accumulated within the neurons with aging. In brains less than 25 years of age there were many fewer pigment-containing neurons than tyrosine hydroxylase-containing neurons; however, by the fifth decade the number of cells identified by the two markers was comparable. From the first to the tenth decade of life there is over a 50% loss of LC neurons: in four cases from "young" individuals (1-28 years of age) there were 21,084 +/- 653 tyrosine hydroxylase immunostained cells (mean +/- standard error of the mean) on one side of the brain; in seven cases from "old" individuals (60-82 years of age) there were 16,502 +/- 921 pigment-containing cells; and in the three cases from the "oldest" individuals (103-104 years of age) there were 9,493 +/- 1,236 pigment-containing neurons. In both the "old" and "oldest" groups, compared to the "young," there was significantly greater loss of rostral cells than caudal cells. These data indicate a systematic loss of cells such that the rostral, forebrain-projecting neurons decrease in number with aging to a greater extent than do the caudal, spinal cord-projecting neurons.(ABSTRACT TRUNCATED AT 250 WORDS)

Computer visualization techniques were used to map the distribution of dopaminergic neurons within midbrain tissue sections from 5 parkinsonian patients and 3 age-matched control subjects. The Parkinsonian brains had over 50% fewer dopaminergic neurons within the midbrain than age-matched normal brains. The cell loss occurred within the combined substantia nigra (dopaminergic nucleus A9) and retrorubral (dopaminergic nucleus A8) areas (greater than 61%) and the ventral tegmental area (dopaminergic nucleus A10) (greater than 42%). The cell loss was greatest within the ventral portion of the substantia nigra zona compacta. The specific pattern of cell loss is very similar to the pattern of cells that project to the striatum (as opposed to cortical and limbic sites) in animal neuroanatomical tracing experiments. These data suggest that Parkinson's disease preferentially destroys midbrain dopaminergic neurons in nuclei A8, A9, and A10, which project to the striatum.