In vitro studies have shown that the Alzheimer's disease-related presenilin
-1 protein can mediate Notch-1 receptor cleavage during signalling. In the
present study, we compared the distribution of presenilin-1 and Notch-1 rec
eptor immunoreactivities in human embryonic CNS tissue during the first tri
mester of development. Our aim was to gain insight into whether these prote
ins are likely to interact functionally during human fetal brain developmen
t. CNS material was obtained from routine abortions, cryosectioned and stud
ied by means of immunohistochemistry with antibodies to presenilin-1 and No
tch-1. At very early stages of embryonic development (four to five gestatio
nal weeks) intensive presenilin-1 immunoreactivity could be seen predominan
tly in neurites in the ventral horn of the spinal cord, where it overlapped
with 200-kDa neurofilament immunoreactivity. Presenilin-1 immunoreactivity
was also seen in neuroblasts of the ventricular zone of the tel- and mesen
cephalon, as well as of the brainstem. Notch-1 receptor appeared in neurona
l and ependymal cells throughout the CNS. Seven- to eight-week CNS tissue s
howed similar patterns of presenilin-1 and Notch-1 receptor expression in t
he spinal cord and cerebral cortex as was seen at five weeks. Both proteins
were localised in the neuroepithelial cell layer lining the ventricles, as
well as in the cortical plate layer, where immunoreactivity was seen in th
e cell bodies. In addition, presenilin-1 immunoreactivity was seen in thin
neurites in the subplate of the developing cortex. At 10 weeks, presenilin-
1 immunoreactivity was reduced in the spinal cord. These results show that,
although presenilin-1 and Notch-1 receptor are localised to the same diffe
rentiating cell populations in the human cerebral cortex, making a direct i
nteraction possible, these proteins are otherwise confined to different neu
rons or neuronal compartments, suggesting a role for presenilin-1 during ea
rly CNS differentiation that does not involve Notch-1 receptor processing.
Double staining for presenilin-1 in the endoplasmic reticulum and presenili
n-1 in the Golgi showed overlap to some extent in investigated CNS regions,
but not in neurites. This suggests that presenilin-1 function during neuro
genesis is not exclusively correlated to protein processing within the endo
plasmic reticulum and Golgi, but that presenilin-1 may also be involved in
other processes, such as axonal and dendritic outgrowth or synaptic formati
on.
In summary, our findings provide supportive evidence that the presenilin-1
protein is involved in the development and maturation of;he human fetal CNS
. The presence of presenilin-1 immunoreactivity in both the cell bodies and
neurites of developing neurons strongly suggests divergent mechanisms of f
unction for presenilin-1 during human brain development. These may include
interactions with any of the Notch receptor proteins, as well as Notch-inde
pendent mechanisms. (C) 2001 IBRO. Published by Elsevier Science Ltd. All r
ights reserved.