Endocrinology Vol. 143, No. 1 84-90
Copyright © 2002 by The Endocrine Society
Melatonin Enters the Cerebrospinal Fluid through the Pineal Recess
Hélène Tricoire,
Alain Locatelli,
Philippe Chemineau and
Benoît Malpaux
Unité Mixte de Recherche Institut National de la Recherche
Agronomique–Centre National de la Recherche Scientifique (6073),
Université de Tours "Physiologie de la Reproduction et des
Comportements," 37380 Nouzilly, France
Address all correspondence and requests for reprints to: Benoît Malpaux, Physiologie de la Reproduction et des Comportements, UMR 6073 Inra-Cnrs-Université de Tours, 37380 Nouzilly, France. E-mail: malpaux{at}tours.inra.fr
 |
Abstract
|
|---|
The pineal recess (PR), a third ventricle (IIIV) evagination
penetrating into the pineal gland, could constitute a site of melatonin
passage to the cerebrospinal fluid (CSF) and explain the high
concentrations of melatonin in this fluid. To test this hypothesis, we
characterized melatonin distribution in the IIIV of sheep by CSF
collection in the ventral part of IIIV (vIIIV) and in PR. At 30
µl/min collection rate, melatonin concentrations were much higher in
PR than in vIIIV (19,934 ± 6,388 vs. 178 ±
70 pg/ml, mean ± SEM, respectively,
P < 0.005), and they increased in vIIIV when CSF
collection stopped in the PR (P < 0.05). At 6
µl/min, levels increased to 1,682 ± 585 pg/ml in vIIIV and were
not influenced by CSF collection in the PR. This concentration
difference between sites and the influence of PR collection on vIIIV
levels suggest that melatonin reaches the PR and then diffuses to the
IIIV. To confirm the role of PR, we demonstrated that its surgical
sealing off decreased IIIV melatonin levels (1,020 ± 305 pg/ml,
compared with 5,984 ± 1,706 and 6,917 ± 1,601 pg/ml in
shams or animals with a failed sealing off, respectively,
P < 0.01) without changes in blood levels.
Therefore, this study identified the localization of the main site of
penetration of melatonin into the CSF, the pineal recess.
|
Introduction
|
|---|
THE CEREBROSPINAL FLUID (CSF), a clear,
colorless liquid, fills the ventricles and external surfaces of the
brain, its total cranial volume being about 60 ml in humans
(1) and 15 ml in sheep (2). The major part of
CSF is secreted by the choroid plexus, a vascular expansion found
mainly in lateral ventricles. Then it circulates from the lateral to
the third and fourth ventricle before entering the subarachnoidal space
and flowing downward around the spinal cord. Several functions were
attributed to the CSF, including buoyancy and protection of the brain,
excretion of metabolites, and homeostasis of the brain chemical
environment (3). More recently, it was suggested that it
could also be an endocrine pathway for intracerebral transport between
different brain areas (1, 3). Indeed, many substances have
been detected in the CSF (4), some of which can penetrate
the cerebral tissue as demonstrated by diffusion studies and by the
physiological and behavioral effects following their
intracerebroventricular injections. However, it is not clear whether
their presence reflects a signal that uses the CSF circulation to
distribute information far away from its release site or a by-product
that spills over into the ventricles after accomplishing its function
in the brain (1). To demonstrate that the presence of a
molecule constitutes a signal in the CSF, several criteria must be met
(1). First, the signaling substance should enter the CSF
and be distributed within the brain by fluid movement. Second, the
signal binding sites leading to the appropriate responses should be
approachable by diffusion or by a specific trapping system. Third, the
removal or replacement of the signaling substance should result in the
correlative response. Considering this, there are already pieces of
evidence for the existence of such signals in the CSF, for example,
diffusible circadian signals delivered from suprachiasmatic
nucleus transplants into the third ventricle (IIIV) that restore
circadian rhythms (5) or IL-1ß-increased levels
accompanying sleep deprivation (6).
Melatonin, the pineal hormone secreted exclusively at night by the
pineal gland, is a good candidate for being transported to its targets
by the CSF for several reasons. First, many of its putative effects
result from a central action, for example, the protection of the
nervous system in neurodegenerative Alzheimer’s or Parkinson diseases
(7), the enhancement of sleep (8), and
immunity (9). Most important, for one well-characterized
effect of this hormone, the seasonal control of LH release, the target
site of melatonin is localized in the premammillary hypothalamic area
(10). In this structure, binding sites are very close to
the CSF of the IIIV (0–1.5 mm), making them easily reachable by
diffusion of melatonin from CSF. Second, melatonin is present in the
ventricular system (11, 12, 13, 14), particularly in the IIIV in
which melatonin concentration is 20 times as high in the CSF as in
blood (15).
One key question is how such a difference in concentrations of the same
molecule between two liquid compartments, CSF vs. blood, can
be obtained. Several hypotheses have been raised to explain the
relative high concentration of CSF melatonin: active uptake of
melatonin from peripheral blood or release from choroid plexus after
retrograde transport from the Galen vein (15, 16, 17, 18).
However, none of these hypotheses has received convincing experimental
support. Interestingly, the pineal gland is intimately related to the
IIIV, of which an evagination, the pineal recess (PR), penetrates into
the organ. This recess, in which the CSF circulates, separates two
laminae forming the stalk and attaching the epiphysis to the brain
(19). According to this, the PR could constitute a site of
melatonin passage to the CSF, either by simple diffusion from pineal
extracellular fluid or direct release of the molecule from CSF
contacting-pinealocytes (20). To test whether pineal
melatonin enters the CSF via the PR, we performed two complementary
experiments. In the first one, we measured in parallel CSF melatonin
levels in the PR and in the ventral part of the third ventricle (vIIIV)
(i.e., a location in which melatonin should be diluted if it
originates from the PR). In the second study, we sealed the pineal
recess to determine whether it prevents melatonin release in the CSF.
The study was performed in sheep because this species enables CSF
collection for a prolonged period without physiological disturbance
(21) and because the anatomical relationship between the
pineal gland and the IIIV is identical in sheep and primates including
humans (19).
|
Materials and Methods
|
|---|
General
Ile-de-France ewes were fed daily with hay, straw, and corn, and
water was available ad libitum. In all experiments, ewes
were restrained so that they were just able to move forward and
backward but were always in contact with other sheep to prevent the
stress of social isolation. Animals were kept in light-controlled rooms
(Exp 1: 12 h of light per day, lights on 0000 h,
lights off 1200 h; Exp 2: 16 h of light, lights on 0600 h,
lights off 2200 h). All procedures were carried out in accordance
with Authorization A37801 of the French Ministry of Agriculture.
Exp 1: does melatonin reach the CSF in the third ventricle directly
from the pineal gland?
CSF melatonin concentrations in the PR and vIIIV were compared
by simultaneous collection in these two sites, and then the influence
of PR collection, the hypothetical source of CSF melatonin, on hormone
levels in the vIIIV was assessed by stopping PR collection. In
addition, this protocol of CSF collection was performed at two
different flow rates to evaluate collection impact on measured
concentrations.
Cannula implantation
Six Ile-de-France ewes were implanted stereotaxically with two
guide-cannulae for CSF collection, as described previously
(15). Briefly, radio opaque material (Omnipaque,
Nycomed Ingenon SA, Paris, France) was injected in the
lateral ventricle through a cannula, delimiting the ventricular system,
particularly the IIIV. The first 40-mm-length x 1.5-mm-diameter
guide cannula was placed to end at the basis of the PR and the second
55-mm-length, in the vIIIV (Fig. 1
).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 1. Schematic representation of the ventricular system
and the implantation of cannulae in Exp 1. I, Cannula in the PR; II,
cannula in the ventral part of IIIV; a, distance between the TMI basis
and PR cannula; b, distance between cannulae I and II. P, Pineal gland;
LV, lateral ventricle; TMI, thalamus massa intermedia.
|
|
Study design
One week later, CSF was collected as previously described
(15) during two consecutive 12-h nights. Each night was
divided into an "experimental period" and a "control period"
randomized among animals according to a cross-over design. The first
one began at least 1 h after the onset of darkness and the last
one finished 1 h before the end of darkness. CSF was collected
throughout the two periods in the vIIIV. In the PR, it was collected
throughout the control period and collection stopped after 2 h of
the experimental period. The CSF fractions were collected every 15 min,
except during the 30-min period following the cessation of CSF
withdrawal in the PR, when 5-min CSF fractions were collected. This
withdrawal design was performed at a flow rate of 30 µl/min during
one night and 6 µl/min during the other night, randomized among
animals according to a cross-over design. The flow rates of CSF used in
this study were below the CSF formation rate in sheep of about 120
µl/min (2, 22). Collection of CSF samples in the PR was
stopped on two animals before the end because the catheter stopped
flowing. These animals were therefore excluded from the study and only
the data of the other four animals were analyzed.
Exp 2: does surgical obliteration of pineal recess prevent
melatonin release in the third ventricle?
The PR of ewes was surgically sealed off to prevent CSF exchange
between the pineal gland and the ventricle. Jugular blood melatonin
concentrations were measured before and after the sealing off as an
index of melatonin synthesis and pineal gland integrity. Three months
later, the nocturnal patterns of CSF and jugular blood melatonin
concentrations were measured for each animal to evaluate sealing-off
effects.
Surgical sealing off of the pineal recess
The PR of 15 ovariectomized ewes with E2 implants was surgically
sealed. After general anesthesia of the ewes (halothane 4%,
Pitman-Moore France, Meaux, France), the scalp was
stripped from midline on the left side of the head and a 50-mm-diameter
hole was made in the skull behind the three-way junction of the
parietal and two frontal bones. The dura mater was opened carefully in
half moon on the left side of the midsagittal sinus. The left
hemisphere was retracted gradually, and adjacent blood vessels were
correctly photocoagulated with diode laser 980 nm (CERALAS D25,
CeramOptec, Bonn, Germany) to expose the pineal gland. The
roof of the IIIV was opened 1–2 mm in front of the pineal gland, above
the PR, using the diode laser. Two drops of biological glue
(Histoacryl, B. Braun Surgical GmbH, Melsungen, Germany)
were placed, sealing the PR and closing the ventricular system. The
dura mater was sutured and the hole in the skull was filled with
acrylic cement. Six sham ewes underwent the same surgical approach
including the pineal exposure but without opening of the IIIV. As a
control for the presence of biological glue in the ventricular system,
a 20-µl piece of prepolymerized glue was introduced in the lateral
ventricle.
Cannula implantation and assessment of sealing-off
effectiveness
About 2 months after the surgical sealing off of the recess, a
guide-cannula was implanted in IIIV, using a lateral x-ray picture.
This picture enabled to see whether the radio opaque fluid penetrates
the PR. Optical density level in the PR and in the area dorsal to the
Sylvius duct used as a black standard for the amount of x-rays crossing
head tissue, was measured on x-ray picture via image analysis framework
(Biocom Histo 500, Les Ulis, France). The absence of radio
opaque in the PR was interpreted as complete sealing off of the PR. In
contrast, the presence of radio opaque in this structure
(i.e., the same distribution as in sham (S) animals) was
interpreted as a failed sealing off (see Fig. 5). Experimental animals
were distributed accordingly into two groups: sealed off (SO) and
failed sealed off (F). At the end of the procedure, a 50-mm
guide-cannula was implanted in the IIIV, allowing the CSF withdrawal,
as described above.
Blood and CSF collections
The effect of surgery on pineal activity was assessed by
measuring blood melatonin levels on average 13 d before and 30,
58, and 124 d after the sealing off (min, max: d -21 to d -4; d
+26 to d +35; d +50 to d +67, and d +116 to d +133, respectively). On
each occasion, four blood samples were obtained by jugular venipuncture
at 2330, 000, 0030, and 0100 h.
On d +119, a polyethylene catheter was inserted into the jugular vein
and secured to the skin of the 18 animals (8 SO, 4 F, and 6 S animals)
to collect 3-ml blood samples, and CSF was collected continuously via a
peristaltic pump. We obtained CSF fractions and blood samples every 30
min, except for one SO animal and one S in which CSF collection was not
successful. The collection period started 1 h before the beginning
of an 8-h night and finished 1 h after its end.
Sample processing and melatonin assay
Plasma and CSF fractions were stored at -20 C until assay.
Plasma melatonin concentrations were determined in 100-µl aliquots,
using a well-validated RIA (19). This assay system was
also used to measure melatonin concentrations in 1- to 50-µl aliquots
of CSF samples, but the standard curve was made in assay buffer instead
of plasma (11). Assay sensitivity averaged 4 pg/ml for
100-µl plasma sample and 40 pg/ml for 10-µl CSF sample. Intra- and
interassay CVs were 7.4% and 11.4%, respectively, for plasma, and
12.4% and 10.9%, respectively, for CSF.
Analysis
In Exp 1, log-transformed CSF concentrations were analyzed by
repeated-measure ANOVA (day of sampling [low vs. fast flow
rate]), period [experimental vs. control], and sample
number as within factors). In Exp 2, blood concentrations on d -13,
+30, +58, and +124 were analyzed by a repeated-measure ANOVA (group as
between factor and day as within factor) after log transformation. On d
+119, blood and CSF concentrations were analyzed by repeated-measure
ANOVA (group and compartment as between factors, time of sampling as
within factor) after log transformation.
|
Results
|
|---|
Exp 1: does melatonin reach the CSF in the third ventricle directly
from the pineal gland?
Melatonin concentrations were much higher in the PR than in the
vIIIV during simultaneous collection at a flow rate of
30 µl/min (178 ± 70
vs. 19,934 ± 6,388 pg/ml, P < 0.001,
ratio PR/IIIV of 182 ± 86, Figs. 2, 3, and 4, mean ±
SEM). Melatonin
concentrations were very variable among
animals (PR from 2,200 to 34,700 pg/ml; IIIV from 74 to 417 pg/ml).
During h 3 and 4 of the experimental period, vIIIV concentrations
increased 63 ± 46-fold when CSF collection stopped in the PR
(from 178 ± 70 to 8,400 ± 3,775 pg/ml, P <
0.05, Figs. 2 and 3), but they did not in control period. However,
those last vIIIV melatonin concentrations (control period: h 3 and 4)
were not significantly different from those measured at the beginning
of the experimental period (h 1 and 2) (296 ± 132 vs.
403 ± 238 pg/ml, experimental vs. control, Figs. 2 and 3). At the lower collection flow rate, concentrations were also much
higher in the PR than in the vIIIV (18,106 ± 5,484 vs.
1,682 ± 585, P < 0.05, Figs. 2 through 4).
However, this difference was smaller than that observed at the higher
flow rate because, with the reduction in flow rate, levels in the vIIIV
increased (178 ± 70 vs. 1,682 ± 585, 30 µl/min
vs. 6 µl/min, respectively, P < 0.05,
Figs. 2 through 4) and were not different in the PR (19,934 ±
6,388 vs. 18,106 ± 5,484 pg/ml; 30 vs. 6
µl/min, respectively, Figs. 2 through 4). The examination of the
x-ray picture revealed a relationship between precise placement of the
cannulae and observed melatonin concentrations at the flow rate of 30
µl/min: the ratio PR/IIIV is positively correlated with the distance
between the thalamic mass and the tip of guide-cannulae located at the
base of the PR (r2 = 0.53; P <
0.05) and negatively correlated with the distance between the two
cannulae (r2 = 0.63; P < 0.05).
These correlations are not significant at a lower collection flow
rate.
View larger version (30K):
[in this window]
[in a new window]
|
Figure 2. Melatonin concentrations in the CSF collected
simultaneously in the pineal recess (+) and in the ventral
part of the third ventricle ( ) of two individuals (top and
bottom), at fast flow rate (30 µl/min) and slow flow rate (6
µl/min).
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Figure 3. Mean ± SEM melatonin levels in
the CSF collected simultaneously in the pineal recess (+)
and in the ventral part of the third ventricle () of four sheep.
Because the combination of phases (experiment vs. control)
and flow rates (30 vs. 6 µl/min) were randomized according
to a cross-over design, the phase order of presentation does not apply
to all animals.
|
|
Exp 2: does surgical sealing off of pineal recess prevent melatonin
release in the third ventricle?
The analysis of eight x-ray pictures fit the criteria of complete
sealing off (absence of radio opaque in PR: optical density similar to
the black standard), and four did not (presence of radio opaque in PR:
optical density similar to that in S and significantly higher than that
of completely SO animals; P < 0.005; Fig. 5
). This observation led to groups of
SO (n = 8) and F animals (n = 4) in addition to S
animals (n = 6).
On d +119, a nighttime increase in blood melatonin concentrations was
observed in all animals regardless of their group. In addition,
nocturnal melatonin levels were not different among groups (338 ±
56, 459 ± 101, 374 ± 69 pg/ml for S, F, and SO animals,
respectively, Figs. 5 and 6). In the CSF,
a nighttime increase in melatonin concentration was also observed in
all animals, but, in contrast to blood, its amplitude varied greatly
among groups. Nocturnal CSF melatonin concentrations were greatly
lower in SO, compared with S and F animals, (ratio SO/S: 15%,
P < 0.05 and SO/F: 17%, P < 0.01;
1020 ± 305, 5984 ± 1706, and 6917 ± 1601 pg/ml in SO,
S, and F, respectively; Figs. 5 and 6).
View larger version (37K):
[in this window]
[in a new window]
|
Figure 6. Effect of the PR sealing off on the melatonin
levels in jugular plasma and in the IIIV CSF. ***,
P < 0.005.
|
|
The changes in melatonin concentrations in the jugular blood, used as
an index of the secretory ability of the pineal gland, were not
different among the three groups of animals following the surgery (no
significant group x time interaction). An effect of time was
found (P < 0.05) with d +30 being different from d +58
(P < 0.05) and no other difference between time points
(d -13: 457 ± 50; d +30: 495 ± 66; d +58: 389 ± 49;
d +124: 409 ± 47 pg/ml).
|
Discussion
|
|---|
In the literature, anatomical observations have suggested several
hypotheses to explain how melatonin could be released in the CSF
[i.e., active uptake of melatonin from peripheral blood or
release from choroid plexus after retrograde transport from the Galen
vein (15, 16, 17, 18)]. Our study performed in sheep, in which
the localization of the pineal gland relative to the IIIV is similar to
that in humans, is the first study aimed at testing one of these
hypotheses. It demonstrates that the major part of CSF melatonin enters
the IIIV through the PR, and from there diffuses to the whole IIIV.
This conclusion is based on three pieces of evidence: 1) CSF nocturnal
melatonin levels are much higher in the PR than in the vIIIV
(i.e., at a distance of less 10 mm); 2) CSF withdrawal in
the PR causes a flow rate-dependent decrease in melatonin
concentrations in the vIIIV; and 3) the surgical sealing off of the PR
causes a dramatic decrease in nocturnal melatonin levels in the
CSF.
The previous observation that melatonin levels were lower in the
lateral ventricle than in the IIIV suggested that melatonin enters the
CSF in the IIIV (15). The present demonstration of higher
melatonin levels in the PR than in the vIIIV extends this conclusion
and suggests strongly that the major access of melatonin to the CSF is
near the PR. The most striking observation in favor of a PR origin is
that the collection of CSF near this site causes a concentration drop
in the vIIIV, and it is dependent on collection flow rate (high at fast
rate, low at low rate). The most obvious interpretation is that
collecting CSF at the source of melatonin deprived the whole
compartment of the hormone; an increase in flow rate leads to a higher
removal of melatonin and therefore a larger drop in its concentrations
downstream, especially in the vIIIV. Alternatively, the high flow rate
of collection (30 µl/min) in the PR is a significant proportion of
CSF production (about 100 µl/min in sheep) (22) and
causes CSF to circulate faster from the rostral to the caudal end of
the IIIV. Therefore, melatonin could not have a chance to progress
rostrally against this stream from the PR to the vIIIV. Regardless of
the explanation, this observation strongly indicates that, in
physiological conditions, CSF melatonin originates from the PR and
diffuses rostrally toward the rest of the ventricle. This implies that
melatonin diffuses from the PR to the vIIIV against the dominant
rostrocaudal flow.
The difference in concentrations between the two sites observed during
simultaneous collection must be taken with great caution in terms of
physiological comparison because it was artificially overestimated by
the collection procedure, particularly at a flow rate of 30 µl/min.
Indeed, concentrations in the PR are not dependent on flow rate and are
likely to reflect closely physiological concentrations in this site.
However, concentrations in the vIIIV are greatly underestimated at a
flow rate of 30 µl/min and, to a lesser extent, at 6 µl/min. In
this latter situation, they agree with previous observations with a
single site measurement at flow rates of 30 or 10 µl/min
(15) and the increase following cessation of PR collection
is much reduced, compared with the other flow rate. An additional
difficulty for a precise assessment of concentrations lies in the
interanimal variability. This variability originates in part in the
well-described differences in melatonin production among animals
(23). This biological variability could also be increased
by the influence of the position of the cannula on the measured levels.
Indeed, a gradient of melatonin concentrations exists between the PR
and the base of the ventricle and a small difference in placement may
cause substantial differences in measured concentrations. Regardless of
this difficulty in quantifying precisely the difference between sites,
the main conclusion that concentrations are higher in the PR than in
the rest of the IIIV remains. Indeed, the concentrations measured in
the PR (from 3,000 to 30,000 pg/ml) have never been found in the vIIIV
in previous studies or during the single-site collection of this study
(on average 1,700 pg/ml in this study and 2,000 pg/ml in Ref.
15).
In Exp 2, the obliteration of the PR allowed to avoid the penetration
of CSF inside the PR as evidenced by the disappearance of this cavity
on the x-ray pictures obtained after injection of radio opaque fluid in
the ventricular system. This obliteration was obtained without
preventing the flow of CSF toward the fourth ventricle, as evidenced by
the presence of radio opaque fluid in the Sylvius duct and the fourth
ventricle. This enabled distinguishing the part of CSF melatonin
originating from the PR and that coming from other parts of the
ventricular system. The dramatic reduction in melatonin concentrations
in the IIIV in SO animals, compared with shams, demonstrates that at
least 80% of IIIV CSF melatonin enter the ventricle through the PR.
One difficulty linked to the surgery was to spare the NPY nerve fibers
in the pineal stalk that have been described in the rat
(24) and that could be involved in the regulation of
melatonin synthesis (25). The sham-operated animals
allowed to remove partly this drawback because their pineal gland was
exposed but not to the stage of opening the IIIV and inserting
biological glue in the ventricular system. However, the F animals are
much better controls because they underwent the same surgery and glue
injection as the completely SO ones with the exception that biological
glue fixed near the PR failed to prevent CSF circulation within the PR.
Their CSF melatonin concentration was as high as in S animals, which
indicates that the reduction of CSF melatonin levels in SO animals was
the consequence of the obliteration and not a nonspecific consequence
of the surgery. Moreover, the changes over time in blood melatonin
concentrations, an index of the synthetic activity of the pineal gland,
did not differ among groups, suggesting that pineal secretion of
melatonin was not disrupted by the obliteration. Therefore, these data
demonstrate that the major part of CSF melatonin comes from the PR.
A small amount of melatonin remains in IIIV CSF of SO animals, compared
with S animals (15%). This remaining presence of melatonin is
suggestive of the existence of another minor source of CSF melatonin.
This hormone is small (232 D) and highly lipophilic, and it can cross
easily the membranes and diffuse from surrounding tissue. It could also
come from the choroid plexus that is possibly connected to the Galen
vein by retrograde blood flow (16). An alternative
explanation for the residual presence of melatonin is that the
obliteration of the recess is not full, despite the absence of evidence
for radio opaque in the recess. Considering the huge concentrations in
the PR, a small communication between the pineal gland and the CSF will
increase melatonin levels in the IIIV. In favor of this latter
possibility, in some animals CSF melatonin levels were down to 70 pg/ml
after obliteration. Regardless of its explanation, the presence of
residual melatonin prevents the use of this model of SO animals to test
the relative importance of the CSF and the blood signals to produce the
biological effects of melatonin because, despite an 85% reduction, CSF
concentrations remain about three times as high as blood concentrations
in sealed off and therefore melatonin is not selectively suppressed in
the CSF compartment.
The present data indicate that the major part of CSF melatonin
originates from the PR in sheep, but how likely is this conclusion to
apply to other species? Although the pineal bodies of different mammals
possess a common developmental origin from the posterodorsal region of
the diencephalic roof, there are variations with respect to pineal
position. The sheep pineal gland is a pea-like organ, belonging to the
proximal type: The bulk of the pineal tissue lies closely related to
the IIIV and is supported by the stalk attaching the pineal body, which
is separated by the PR (19). The human pineal gland is a
piriform organ and presents the same anatomical organization as in
sheep, suggesting that our data on CSF melatonin origin are likely to
apply to humans. In contrast, the pineal gland of rodents extends from
the IIIV ("deep" pineal) to immediately beneath the skull
("superficial" pineal) (19). Electron microscopy
investigations in Mongolian gerbil and in the vole (20)
demonstrated that the deep pineal gland-CSF interface is not covered by
a typical ependyma usually protecting the ventricular walls and that
some pinealocytes are protruded in the PR, being in direct contact with
the CSF. These bulging pinealocytes of the deep pineal gland could
therefore release melatonin directly into the PR, and the superficial
pineal could release melatonin preferentially in blood. It will be of
interest to determine whether species with a proximal type pineal, like
humans and sheep, also display bulging pinealocytes, which would
constitute an anatomical support for the melatonin release in the PR
described in this study.
In conclusion, these results provide, for the first time, the
localization of melatonin entry in the CSF. At least for its major
part, melatonin is released into the PR and then is distributed
throughout the ventricular system to reach possibly
periventricular-binding sites. Further investigation is required to
first determine the anatomical support for a direct secretion in the PR
and then the functional role of CSF melatonin.
|
Acknowledgments
|
|---|
We wish to thank J.C. Thiéry for valuable comments
on the manuscript; D. Skinner for helpful discussion; A. Daveau,
O. Léon, and Y. Locatelli for technical assistance; F.
Paulmier, F. Dupont, and their technical staff for animal management;
and M. Charker for correcting the English manuscript. We are grateful
to CeramOptec (Bonn, Germany) for generously providing the diode
laser equipment (CERALAS D25).
|
Footnotes
|
|---|
H.T. was supported by a Ph.D. grant from "Région Center."
Abbreviations: CSF, Cerebrospinal fluid; F, failed sealed-off
animals; IIIV, third ventricle; PR, pineal recess; S, sham animals; SO,
sealed-off animals; vIIIV, ventral part of IIIV.
Received August 15, 2001.
Accepted for publication September 19, 2001.
|
References
|
|---|
-
Nicholson C 1999 Signals that go with the
flow. Trends Neurosci 22:143–145[CrossRef][Medline]
-
Evans CA, Reynolds JM, Reynolds ML, Saunders NR, Segal
MB 1974 The development of a blood-brain barrier mechanism in
foetal sheep. J Physiol 238:371–386[Abstract/Free Full Text]
-
Nilsson C, Lindvall-Axelsson M, Owman C 1992 Neuroendocrine regulatory mechanisms in the choroid
plexus-cerebrospinal fluid system. Brain Res Brain Res Rev 17:109–138[CrossRef][Medline]
-
Wood JH 1982 Neuroendocrinology of cerebrospinal
fluid: peptides, steroids, and other hormones. Neurosurgery 11:293–305[Medline]
-
Silver R, LeSauter J, Tresco PA, Lehman MN 1996 A
diffusible coupling signal from the transplanted suprachiasmatic
nucleus controlling circadian locomotor rhythms. Nature 382:810–813[CrossRef][Medline]
-
Fang J, Wang Y, Krueger JM 1998 Effects of
interleukin-1 beta on sleep are mediated by the type I receptor.
Am J Physiol 274(3 Pt 2):R655–R660
-
Daniels WM, van Rensburg SJ, van Zyl JM, Taljaard
JJ 1998 Melatonin prevents beta-amyloid-induced lipid
peroxidation. J Pineal Res 24:78–82[Medline]
-
Arendt J, Skene DJ, Middleton B, Lockley S, Deacon
S 1997 Efficacy of melatonin treatment in jet lag, shift work and
blindness. J Biol Rhythms 12:604–617
-
Liebmann PM, Wolfler A, Felsner P, Hofer D,
Schauenstein K 1997 Melatonin and the immune system. Int Arch
Allergy Immunol 112:203–211[Medline]
-
Malpaux B, Daveau A, Maurice-Mandon F, Duarte G,
Chemineau P 1998 Evidence that melatonin acts in the premammillary
hypothalamic area to control reproduction in the ewe: presence of
binding sites and stimulation of luteinizing hormone secretion by
in situ microimplant delivery. Endocrinology 139:1508–1516[Abstract/Free Full Text]
-
Rollag MD, Morgan RJ, Niswender GD 1978 Route of
melatonin secretion in sheep. Endocrinology 102:1–7[Abstract/Free Full Text]
-
Shaw PF, Kennaway DJ, Seamark RF 1989 Evidence of high concentrations of melatonin in lateral ventricular
cerebrospinal fluid of sheep. J Pineal Res 6:201–208[Medline]
-
Kanematsu N, Mori Y, Hayashi S, Hoshino K 1989 Presence of a distinct 24-hour melatonin rhythm in the ventricular
cerebrospinal fluid of the goat. J Pineal Res 7:143–152[Medline]
-
Reppert SM, Perlow MJ, Tamarkin L, Klein DC 1979 A
diurnal melatonin rhythm in primate cerebrospinal fluid. Endocrinology 104:295–301[Abstract/Free Full Text]
-
Skinner DC, Malpaux B 1999 High melatonin
concentrations in third ventricular cerebrospinal fluid are not due to
Galen vein blood recirculating through the choroid plexus.
Endocrinology 140:4399–4405[Abstract/Free Full Text]
-
Smulders AP, Wright EM 1980 Role of choroid plexus
in transport of melatonin between blood and brain. Brain Res 191:555–558[Medline]
-
Maurizi CP 1991 Recirculation of cerebrospinal
fluid through the tela choroideae is why high levels of melatonin can
be found in the lateral ventricles. Med Hypotheses 35:154–158[CrossRef][Medline]
-
Quay WB 1973 Retrograde perfusions of the pineal
region and the question of pineal vascular routes to brain and choroid
plexuses. Am J Anat 137:387–401[CrossRef][Medline]
-
Vollrath L 1981 The pineal organ. In: Oksche A,
Vollrath L, eds. Handbuch der mikroskopischen anatomie des menchen.
Berlin: Springer-Verlag; VI/7
-
Hewing M 1982 Pinealocytes contacting the
cerebrospinal fluid of the suprapineal recess in the Mongolian gerbil
(Meriones unguiculatus). Cell Tissue Res 222:177–185[Medline]
-
Skinner DC, Malpaux B, Delaleu B, Caraty A 1995 Luteinizing hormone (LH)-releasing hormone in third ventricular
cerebrospinal fluid of the ewe: correlation with LH pulses and the LH
surge. Endocrinology 136:3230–3237[Abstract]
-
Chodobski A, Szmydynger-Chodobska J, Cooper E, McKinley
MJ 1992 Atrial natriuretic peptide does not alter cerebrospinal
fluid formation in sheep. Am J Physiol 262(5 Pt 2):R860–R864
-
Zarazaga LA, Malpaux B, Guillaume D, Bodin L, Chemineau
P 1998 Genetic variability in melatonin concentrations in ewes
originates in its synthesis, not in its catabolism. Am J Physiol
274(6 Pt 1):E1086–E1090
-
Mikkelsen JD, Moller M 1999 Neuropeptide Y in the
mammalian pineal gland. Microsc Res Tech 46:239–256[CrossRef][Medline]
-
Ribelayga C, Pevet P, Simonneaux V 1998 Possible
involvement of neuropeptide Y in the seasonal control of
hydroxyindole-O-methyltransferase activity in the pineal
gland of the European hamster (Cricetus cricetus). Brain Res 801:137–142[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
V. Simonneaux and C. Ribelayga
Generation of the Melatonin Endocrine Message in Mammals: A Review of the Complex Regulation of Melatonin Synthesis by Norepinephrine, Peptides, and Other Pineal Transmitters
Pharmacol. Rev.,
June 1, 2003;
55(2):
325 - 395.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
) during an 8-h
night, in an S, an F, and an SO animal.
) or low (
) flow rate
collection of CSF on melatonin levels (mean ± SEM) in
the ventral part of the IIIV and in the PR. *, P <
0.05; **, P < 0.001.