v

 

 

 

MORPHOLOGICAL CHANGES IN HYPOTHALAM‘IC
NEUROSECRETORY NUCLEI DURING THE DIURNAL CYCLE

Thesis for the Decree of M. A

MICHIGAN STATE UNIVERSHY

WILLIAM EARL ARMSTRONG
1977

:7'

- 1F

 

 

'71, AI ‘ ‘ ‘
an .
‘1..- _.'. -. .
. 1|
" J! .
£11.}. ~ --
. v
o ‘- ‘1 - ‘
“ I . g -,'V -

. '."J' A -“. ;
5' e?
'mmm‘wn'

M

. ' . O
.V‘ . 1 P 14'
.

t

I

. A“

.4. 2r.

'3

5.363311; I. . _

 

 

 

 

 

ABSTRACT
MDRPHOLOGICAL CHANGES IN HYPOTHALAMIC

NEURDSECRETORY NUCLEI DURING THE
DIURNAL CYCLE

BY

William Earl Armstrong

At different times during a 12:12 light—dark cycle, the occurrence of
cells with multiple and marginated nucleoli and cell area changes were ex-
amined in hypothalamic neurosecretory perikarya known to contain either
vasopressin (V?) or oxytocin (OX) in the albino rat. These measures were
used as evidence for neurosecretory activity in the suprachiasmatic (SCN),
supraoptic (80a) and paraventricular (PVN)nuclei, and were compared with
plasma corticosterone values in order to partially assess the relationship
of these nuclei with ACTH regulation. The relationship of these nuclei with
neurohypophysial function was gleaned from comparing the morphological re-
sponses measured herein with previous descriptions of plasma and pituitary
VP.

Iluctuations in plasma VP during the day may underlie urinary excretory
rhythms. Rats excrete more urine during the night than during the day.
Accordingly, 80a neurons exhibited their highest percentage of cells with
multiple nucleoli at the end of the day, indicating this nucleus had been

‘more active when lights were on. The periodicity of this response was in

William E. Armstrong

phase with that of plasma corticosterone concentrations. The changes in
this nucleus make a role in both neurohypophysial and adenohypophysial
'function feasible.

The release of ACTH from the adenohypophysis may be partly regulated
by the VP found in the external layer of the median eminence (MEx). It
was hypothesized that the source of this VP may be the dorsomedial portion
of the SCN (SCdm), shown to contain VP—positive cell bodies using immuno-
cytochemical techniques. However, the percentage of cells with multiple
nucleoli did not change in the SCdm, or in the ventrolateral SCN (SCvl)
during the day. Cell sizes in bg£h_nuclei appeared to vary slightly, with
a peak about lights off. This response may represent endocrine activity,
or general neuronal activity related to circadian rhythms.

Evidence was provided which indicated there are morphological
differences between SCdm and SCvl. Cells in SCvl were larger and more
often had an invaginated nucleus than did those in SCdm.

Both lateral (PVl) and medial (PVm) magnocellular divisions of the
PVN showed a large increase in cell areas which peaked at 2400 hr in the
absence of any fluctuations in the percentage of cells with multiple
nucleoli. Recent evidence showing complex extrahypophysial connections
of the PVN makes this response difficult to interpret. A role in ACTH
regulation for the VP in the PVN has been suggested by research using
monkeys. It appears that OX and VP cells were simultaneously active in
the PVN, and that these hormones may have central nervous system roles in
addition to their known secretion from the neurohypophysis. A dissociation
between the 80a and PVN was provided by differences in the nucleolar and
areal responses of the two nuclei.

As to the existence of multiple nucleoli, there were two novel

findings. Firstly, previous studies on the SOa had shown that, once

William E. Armstrong

nucleolar proliferation was induced, the level of cells with more than one
nucleolus remained high for several hours, and even days, after the activat-
ing stimuli were terminated. The present study indicates nucleoli may be
produced, and subsequently removed from the nucleus in approximately six
hours during the diurnal cycle. Secondly, cells with multiple nucleoli
were found to be prevalent in the SCN, though their significance in this

nucleus remains to be determined.

MORPHOLOGICAL CHANGES IN HYPOTHALAMIC
NEUROSECRETORY NUCLEI DURING THE
DIURNAL CYCLE

By

William Earl Armstrong

A thesis
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF ARTS
Department of Psychology

1977

To My Wife, Dee Anne, And
My Parents, Mr. and Mrs.

Earl I. Armstrong

11

ACKNOWLEDGEMENTS

Dr. Glenn I. Hatton was particularly supportive and patient during the
development of this thesis. Drs. Charles D. Tweedle and John I. Johnson pro-
vided critical suggestions on the manuscript while serving on my committee.

Undying moral support came from my wife, Dee Anne, and my parents, Mr. and

Mrs. Earl Armstrong.

This research was supported by NIH grant No. N309140 to Dr. Glenn I.

Hatton.

iii

TABLE OF CONTENTS

DEDICATION .......................................................... 11
ACKNOWLEDGEMENTS .................................................... 111
LIST OF TABLES ...................................................... v
LIST OF FIGURES ..................................................... vi
LIST OF APPENDICES ..................................................viii
INTRODUCTION ........................................................ 1
METHOD .............................................................. 8
mwmms.H.H.H.H.H.H.H.H.H.n.u.u.u.u.n.u.n.u.u.u. 15
DISCUSSION .......................................................... 36
LIST OF REFERENCES .................................................. 47
APPENDIX A (Materials)............................................... 53
APPENDIX B (Histology)............................................... 54

APPENDISC (Raw Data)00000000000OOOOOOOOOOGOOOOOOOOOOOOOOOOOOOOOOOOO 56

iv

LIST OF TABLES

Table Page

1. Plasma protein and hematocri£_values at
different times of the day (x:SEM).....................16

2. Mean (:_SEM) percentage of cells with
marginated nucleoli in SCdm and SCvl
at different times Of the day...00.0.00...0.0.0.00000000023

3. Mean (1 SEM) percentage of cells with
marginated nucleoli in the SON and PVN
at different times of the day............................28

LIST OF FIGURES

Figure

l.

Photomicrograph of a 22nm coronal section through the suprachiasmatic
nuclei, illustrating dorsomedial (dm) and ventrolateral (v1) divisions.
The brain was embedded in celloidin and stained with thionine. CO:
optic chiasm, V: third ventricle.

Photomicrograph at high magnification illustrating cell types in the
suprachiasmatic nucleus. Large arrow indicates a cell with an invaginated
nucleus; small arrows point to two nucleoli in another cell. The brain
was embedded in plastic (methyl methacrylate), sectioned at 2pm, and
stained with thionine.

In the top panel the mean percentages of cells with multiple nucleoli
are graphed for the supraoptic nucleus (SOa) as a function of time of
day. Mean corticosterone values (in ug/100 m1 of plasma) from cardiac
blood are similarly plotted in the bottom panel. In this and in all
other figures, the dark bar indicates the period of darkness. Note
the peak values at 1800 hr, and the troughs at 2400 and 0600 hrs.

The mean percentages of cells exhibiting a nuclear invagination in the
dorsomedial (SCdm) and ventrolateral (SCvl) portions of the supra-
chiasmatic nucleus are plotted as a function of time. SCvl exhibits
significantly higher values at all time, except at 0600, where con-
vergence can be observed.

In the top panel the mean percentages of 'multiples' for all sampled
cells in either the dorsomedial (SCdm) or ventrolateral (SCvl) portion
of the suprachiasmatic nucleus are graphed as a function of time of day.
Depicted in the middle panel are similar values for cells exhibiting

an invaginated nucleus. In the bottom panel the mean percentages for
cells without a nuclear invagination are also plotted. Cells with
invaginated nuclei exhibit some variation in their percentage of'multi-
ples', with a peak.atL1800 hr. Cells without invaginated nuclei exhibit
no such variation.

Depicted are mean cell areas (in umz) in the dorsomedial (SCdm) (top
panel) and ventrolateral (SCvl) (bottom panel) portions of the supra-
chiasmatic nucleus as a function of time of day, represented by solid
line (scale t the left). Also shown are the mean values of difference
scores (in um ) from each subnucleus obtained by subtracting the mean
cell area of each plane from each animal's mean value, represented by
the broken line (scale to the right). Cell areas peak at 1800 hr for
both subnuclei. The changes in cell area appear much more profound
when the differences between planes are minimized, as can be seen by
examining the difference scores. H: cells taken from.horizontal plane
C: cells taken from coronal plane.

vi

10.

Photomicraph of a 6 pm coronal section through the paraventricular
nucleus, illustrating magnocellular neurosecretory neurons, some of
which contain multiple nucleoli (large arrows). Small arrows point
to marginated nucleoli. The brain was embedded in paraffin and
stained with thionine.

Mean cell areas (in umz) from the supraoptic nucleus (SOa) (top panel)
and hippocampus (HPC) (bottom panel) are plotted against time of day,
and represented by solid line (scale to the left). Also shown are the
mean difference scores (in pm ) from each nucleus, obtained by sub-
tracting the mean cell area of each plane from each animal's mean value,
represented by broken lines (scale to the right). In the top panel,
note the separate peaks in area for cells in the horizontal (2400 hr)
and coronal (1800 hr) planes. Total cell area, as judged by difference
scores, peaks at 1800 hr. Note magnitude of the difference scores

at 1800 and 2400 hrs. In the bottom panel, HPC cell areas peak at

1800 hr, and quickly drop off at 2400 hr. These changes are similarly
reflected in the difference scores. H: cells taken from the horizontal
plane C: cells taken from the coronal plane.

The mean percentages of cells with multiple nucleoli are plotted for

the lateral (PVl) and medial (PVm) divisions of the paraventricular
nucleus as a function of time of day. The PVl shows a higher percentage
of 'multiples' at all times than does the PVm. Neither the PVl or

the PVm vary in their percentages of 'multiples' significantly over
time.

Mean cell areas (in umz) from the medial (PVm) (top panel) and lateral
(PVl) (bottom panel) divisions of the paraventricular nucleus are
graphed as a function of time of day, and represented by solid lines
(scale to the left). Also shown are mean difference scores obtained by
subtracting the mean cell area of each plane from each animal's mean
value, represented by broken lines (scale to the right). The peaks

in cell area in both the PVl and the PVm occur at 2400 hr. This change
is punctuated further by a similar change in the mean difference scores.
H: cells taken from the horizontal plane C: cells taken from the
coronal plane.

vii

LIST OF APPENDICES
Page
Appendix A: Materials.............................................53
Appendix B: Histology.............................................54

Appendix C: RawDataOOO0.0ICOOOOOOOOOOOOOOOOO0.0.0.0.000000000000056

viii

INTRODUCTION

Diurnal oscillations characterize a broad spectrum of an organism's be-
havioral and physiological activities. Rhythmic endocrine activity is of
particular interest because of the widespread effects that hormones exert
on regulatory and other adaptive functions. Perhaps the best documented of
established vertebrate hormonal rhythms are those of plasma corticosteroids,
occurring in species ranging from goldfish (Singley and Chavin, 1971) to
man (Migeon et al., 1956). Most intensively studied are rats, which exhibit
circadian variations in all aspects of the pituitary-adrenal axis, including
adrenal (Critchlow, 1963) and plasma (Critchlow et al., 1963; Guillemin et
al., 1959; Retienne et al., 1968) corticosterone, plasma and pituitary ad-
renocorticotropin hormone (ACTH) (Critchlow, 1963; Retienne et al., 1968)
and from median eminence extracts, corticotropin-releasing factor (CRF)
(Hiroshige, Sakakura and Itch, 1969). The persistent periodicity in CRF
secretion in adrenalectomized (Hiroshige and Sakakura, 1971) and hypophy-
sectomized (Takebe, Sakakura and Mashimo, 1972) rats attests to the rhythm's
neural basis and relative independence of negative feedback exerted by ACTH
or corticosterone.

Despite the fact that CRE was the first demonstrable hypothalamic re-
leasing factor, attempts at purification have been largely fruitless (see
Blackwell and Guillemin, 1973). Schally et a1. (1960) did isolate two small
polypeptides from neurohypophysial extracts, beta and alpha CRF, and noted their
similarities to lysine vasopressin and alpha melanocyte-stimulating hormone,
respectively. While this argues for a biochemical separation between vaso-

pressin (VP; the principal mammalian antidiuretic hormone) and CRF, similarities

2

in physiological functions have been demonstrated and were reviewed by McCann
and Dhariwal (1967) and Yates and Maran (1974). Exemplary among these are the
common release of VP and ACTH under many of the same, often stressful, con-
ditions, and the ability of VP to release ACTH when injected into the adeno-
hypophysis of the conscious dog.

Relationship of VP with CRF

The interpretations of studies directed at dissociating CRF and VP re—
lease to certain stimuli are often confounded by investigators' persistence
in drawing conclusions about adenohypophysial VP from its concentration in
peripheral plasma. For example, Hedge and De Wied (1971) report that hypoth-
alamic atropine implants inhibit the stress-induced release of ACTH in the
rat, but not the rise in plasma VP. Indeed, Zimmerman et a1. (1973) were
able to measure VP in the long portal vessels of the anesthetized monkey
using a sensitive radioimmunoassay, and found concentrations §QQ_times that
obtained concurrently from peripheral blood.

Yates and Maran (1974) also argue that neurohypophysectomized rats with
normal ACTH-releasing capabilities may be used as evidence for a separation
of VP and CRF. While presumably eliminating the major source of peripheral
VP, neurohypophysectomy may not affect concentrations in the pituitary portal
system. Terminals from the supraoptic (SON) and paraventricular (PVN) nuclei
may exist rostral to the neurohypophysisalong the median eminence (see Parry
and Livett, 1973). Both of these nuclei synthesize VP and oxytocin (0X)
(Burford et al., 1974). There is strong evidence, however, that VP is not
the sole CRF, but participates in some conjunctive manner to regulate ACTH
secretion. Rats homozygous for hereditary diabetes insipidus, and therefore
lacking the ability to synthesize VP (Valtin, 1967), exhibit ACTH release
when stressed (see MCCann and Dhariwal, 1967, and Yates and Maran, 1974),

although this release may be impaired (Wiley et al., 1974).

3
Recent findings have favored the hypothesis that VP, or some derivative

thereof, may act in concert with another substance to release ACTH (Yates and
Maran, 1974). Yasuda and Greer (1976) have found that neither arginine or
lysine vasopressin is effective in stimulating ACTH secretion from cultured
adenohypOphysial cells, whereas Pitressin (posterior pituitary extract) in—
duces a dose dependent release. Using gel filtration, two distinct peaks
of CRF-active substances from rat median eminence extracts can be separated
(Pearlmutter et al., 1975). A combination of the two fractions releases more
ACTH in gi££2_than either does alone. Furthermore, derivatives of the lysine
vasopressin molecule high in ACTHrstimulating activity have been demonstrated
in the rat (Arimura et al., 1969). Perhaps VP is transformed ig_gigg_into
other biologically active peptides, one of which may be ofa potent CRF.
Localization of VP

Information regarding the origin and distribution of hypothalamic
hormones has increased tremendously with immunohistochemical data (Zimmer-
man, 1976). Unfortunately, the inability to identify the amino acid sequence
of CRF has in this case precluded the application of these techniques. Imm-
unohistochemical studies have confirmed the presence of VP and OK in both the
SON and PVN of several mammals (Zimmerman, 1976). In addition, and unexpectedly,
the dorsomedial portions of the rat suprachiasmatic nuclei (SCdm) also stain
positive for VP using either immunofluorescent (Burlett and Marchetti, 1975;
Swaab et al., 1975) or immunoperoxidase (Vandesande et al., 1975) techniques.
In light of these findings, it is intriguing that Lang and his co~workers
(1976) have recently found that those nuclei in the rat with significant CRF-
activity correspond with nuclei containing VP. This includes the rat arcuate
nuclei, shown by George and Jacobowitz (1975) to contain VP when tissue homo—
senates were analyzed by radioimmunoassay. Immunohistochemical studies have

failed to localize VP in this nucleus in the rat, however (Zimmerman, 1976).

4
It is at this point not unreasonable to speculate that any of the

hypothalamic nuclei mentioned above might contribute to the VP, or its
associated protein neurophysin, present in the external layer of the rat
median eminence (MEx) (see Zimmerman, 1976). This finding obviously gives
new impetus to the hypothesis that VP participates in adenohypOphysial func-
tion. Increases in the amount of neurOphysin (Vandesande et al., 1974) and
VP (Vandesande et al., 1975) in the rat MEx have been reported following
adrenalectomy. In this situation hypersecretion of CRF might be expected

in the absence of corticosteroid feedback. A similar phenomenon is thought
to occur in the rat SCN (Burlett and Marchetti, 1975; Dierickx et al., 1976;
Vandesande et al., 1974) but is still controversial (Zimmerman, 1976). In
the monkey, it appears the PVN may be a major source of VP in the MEx (Antunes
et al., 1976). Several anatomical characteristics of the SCN and recent phy-
siological reports make this structure a putative source of diurnal fluctua-
tions in CRF in the rat.

SCN and Diurnal Rhythms

 

NUclear groups controlling CRF secretion must receive, or themselves
contain, some source of oscillation which is entrainable to the imposing
light-dark (LD) schedule. This point is well illustrated in experiments
where rats are subjected to a sudden alteration in the LD sequence, or
blinded. Reversing the LD schedule induces a phase reversal of the cortic-
osterone rhythm (Critchlow et al., 1963) whereas blinding (Moore and Eichler,
1972) and constant darkness (Critchlow, 1963) produce phase shifts in the
rhythm.without disrupting its circadian periodicity. The SCN receives a
bilateral retinal projection in the rat (Moore and Lena, 1972) and is nec-
essary for the entrainment and generation of several rhythmic functions
(Moore and Eichler, 1976), including that displayed by corticosterone secretion
(Moore and Eichler, 1972).

5
A retinohypothalamic projection to the SCN has been found in several

vertebrates studied with autoradiographic techniques, including birds (Meier,
1973) and primates (Tigges and O'Steen, 1974) as well as many other mammals
(Moore, 1973). Cobalt-filled profiles of small caliber axons have been shown
entering the ventral portion of the most posterior aspect of the rat SCN
(Mason and Lincoln, 1976). Stimulation of the optic nerve or light excites
approximately half of sampled SCN units, while inhibiting a smaller percent-
age (Nashino et al., 1976). These results give sufficient support to the
notion that SCN normally receives information regarding photoperiod. Blinded
rats, which show a corticosterone rhythm, would still have intact terminals
in the SCN from serotonergic neurons in the brain stem, probably originating
in the dorsal and median raphe' nuclei (Fuxe, 1965; Aghajanian et al., 1969).
This system may provide periodic input to the SCN which is coupled with that
from the eye under normal conditions.

If the SCN is one source of VP found in the MEx, then a connection be-
tween the two structures must exist. Studying the rat, Swanson and Cowan (1975)
observed silver grains over fibers projecting to the periventricular region
of the hypothalamus as well as to an area between the arcuate and ventromedial
nuclei after injecting tritiated leucine and proline into SCN. A small
contingent of fibers was reliably present in the internal layer of the median
eminence, commonly associated with axons from.magnocellular neurosecretory
nuclei. Terminal degeneration has been reported in the rat MEx subsequent
to lesions in the SCN, but not immediately rostral to it (Koritsanszky and
Koves, 1976). Ultrastructurally, SCN morphology is compatible with neuro—
secretion, as dense core vesicles of the 1000-1500 3 range are present in
rats (Subaro and Pellegrino de Iraldi, 1969), and rabbits (Clattenburg et

at., 1975).

6

The existence of diurnal neurosecretory rhythms in rats has been
suggested on the basis of Gormori-positive staining and somal or nuclear size
changes by Rinne and Sonninen (1964), studying the SON and PVN, and by Glantz
(1967), studying the SON. Further, daily variations have been reported in the
concentration of VP in the neurohypophysis (Retienne, et al., 1968) and the
blood (Mfihberg and Mahberg, 1976) which may provide an explanation for some
of the cellular changes and for periodicities in urine concentration and
excretion (Glantz, 1967).

Statement of problem

While diurnal rhythms in the secretion of VP and CRF are well docu-
mented, the hypothalamic source(s) of such hormonal oscillations is un—
known. This is due to the presence of VP in several discrete nuclei, all
of which possess ACTH-releasing activity. The problem, then, is to find
the hypothalamic neurosecretory cells whose oscillatory activity (based
on anatomical changes) would account for known diurnal changes in VP and
CRF in the blood or pituitary gland. Furthermore, would such changes, if
they occurred, provide support for the hypothesis that VP participates in
diurnal ACTH secretion?

In this regard, the problem requires the cellular response to be re—
lated to hormone output. Since corticosterone is easy to measure, and its
level in peripheral blood depends on ACTH levels, which depend on CRF
secretion from the hypothalamus, this hormone was chosen for comparisons
with neurosecretory cell changes. '

Magnocellular neurosecretory nuclei contain a proportion of neurons
with multiple nucleoli which is dependent upon their activity (Hatton and
Walters, 1973; Hoblitzell et al., 1976). Although parvocellular in nature,
SCN cell nuclei may also contain more than one nucleolus (Le Beaux, 1971).

In an attempt to discern possible secretions of VP from SCN, SON, and PVN

7
during the diurnal cycle, this nucleolar response, in addition to cell size,

was used as an index of activity. Comparisons were made among these VP-
containing nuclei in an effort to more precisely determine the origin of
neurohypophysial and peripheral vacillations in VP. To quantify neuro~
secretory activity, the percentage of cells with multiple nucleoli and cell
areas were calculated in SCN, SON, and PVN in four groups of four rats each,
sacrificed at six hour intervals after being on a 12:12 LD schedule for at
least 30 days. For correlational purposes, cardiac blood samples were taken
and analyzed for corticosterone, protein concentration, and hematocrit.

The rhythm of morphological changes in each nucleus was compared to that of
corticosterone. Also, the tOpographic distribution of VP and 0X within the

PVN, and of VP within the SCN, was taken into consideration.

METHOD

Subjects

Subjects were 16 male Sprague-Dawley rats (Holtzman Co.), 60 days old
and weighing 200-300 g on arrival. The animals were individually housed in
wire mesh cages which were placed on a rack and contained in a room under
an alternating (12:12) light-dark (LD) cycle. Fluorescent overhead lights
provided approximately 6-8 ft.-c. of illumination, as measured by a photo-
meter (American Optical) directed toward the faces of the cages. Ambient
temperature ranged from 21-230 C. All rats were provided with food (wayne
Mbuse Breeder Blox) and tap water ad libitum throughout the duration of
the experiment.

Procedure

Rats were assigned to one of four groups, four animals per group:

1) animals to be killed at lights off (1800); 2) animals to be killed six
hours into the dark phase (2400); 3) animals to be killed at lights on
(0600); and 4) animals to be killed six hours into the light phase (1200).
Animals were allowed to remain on the LD schedule for at least 30 days,

but no longer than 32 days, before sacrifice. Groups 1800 and 0600 were
run simultaneously as were groups 1200 and 2400. These latter two groups
were run six months after groups 1800 and 0600.

Blood sampling. For five days prior to sacrifice, the animal room
was entered at the predesignated times and the rats were briefly handled.
This was done in order to attenuate stress before blood sampling and per-
fusion. On the sixth day, each animal was treated in the following way at
its scheduled time. The rat was quickly anesthetized with ether and sub-

8

9

jected to cardiac puncture with a heparinized glass syringe. A 1-2 cc

sample of blood was withdrawn and injected into a glass centrifuge tube.

A small amount of this blood was collected in a heparinized glass capillary
tube, and both samples were centrifuged. Plasma was drawn off the large
sample with a 1 cc tuberculin syringe and a small amount immediately analyzed
for protein concentration via refractometry (American Optical). The rest

of the plasma was frozen for subsequent fluorometric determinations of
corticosterone according to the procedure of Guillemin et al., (1959).1
Hematocrit readings were taken from the capillary tube and expressed as the
percentage of plasma.

Histology. While each rat's samples were being centrifuged, trans-
cardial perfusion was begun with a 0.9% saline solution followed by a
solution of one part 37% formaldehyde and nine parts 0.9% saline. With
the perfusion complete, another rat was taken and the procedure repeated.
Either all four rats from one group or two per group were sacrificed on
the same day. In the latter case the remaining rats were handled but not
killed until the next day. All rats were sacrificed within the hour follow-
ing the designated time. The heads of perfused rats were removed and placed
in the fixative for a few days. The brains were removed and further soaked
in fixative, dehydrated in graded alcohols and methyl benzoate, and em?
bedded in paraffin (Paraplast). Sections (6 pm) were taken on a rotary micro—
tome, floated on warm (52-550C) water, and mounted on gelatinized glass
slides. Following deparaffinization with xylene, the sections were stained
for Nissl substance with thionine. Two brains in each group contributed
horizontal sections, two coronal sections. The slides were coded by an
associate and data were collected blind.

1Thanks are due Dr. Ken.Moore for corticosterone analysis.

10

Nuclei examined. Data were collected from the anterior portion of

 

the supra0ptic nucleus (30a), the paraventricular (PVN) and suprachiasmatic
nuclei (SNC), and from region CA 4 (Lorente de No', 1934) of the hippocampus
(HPC). The PVN was subdivided into lateral (PVl) and medial (Pvm) magno-
cellular divisions according to Hatton et al., (1976). The PVm lies in the
anteroventromedial aspect of the PVN, while PVl is situated in the dorsola-
teral protion of the complex, caudal to PVm. Magnocellular neurons of the
SOa lie dorsolateral to the optic chiasm. Parvocellular neurons of the SCN
were also further subdived into dorsomedial (SCdm) and ventrolateral (SCvl)
portions (see Figure l). The presence of a few large blood vessels and

an apparently greater cell density served to distinguish SCdm from SCvl.
The pyramidal neurons in division CA 4 of the HPC are found in the region
of the dentate hilus.

Cell measurements. For cell size determinations, 60 cells from each

 

subnucleus, in each animal, were magnified 2180 times and their perikaryal
outlines traced. Only cells with a visible nucleolus were drawn. For each
subnucleus, each side of the brain contributed half the total number of cells
drawn. In every case, samples were taken from four sections through a par-
ticular subnucleus, with each section providing 15 cells. Areal measure-
ments were made by drawing perpendicular lines through the long and short

axes of the cell drawing. A weighted average of their lengths

(longaxis + 2x short axis)
2

was then determined. Areas measured by planimetry were obtained from random
samples of 80 cells per plane for each subnucleus, collecting these cells from
several animals to account for changes in shape and size. A productdmoment
correlation co—effecient was then computed between the measured areas and

the weighted average of the diameters, for each group of 80 cells. A re-

gression equation was finally used to obtain comparable measurements on cells

Figure 1.

Figure 2.

ll

Photomicrograph of a 22pm coronal section through the
suprachiasmatic nuclei, illustrating dorsomedial (dm)
and ventrolateral (v1) divisions. The brain was em—
bedded in celloidin and stained with thionine. CO:
optic chiasm, V: third ventricle.

Photomicropgraph at high magnification illustrating
cell types in the suprachiasmatic nucleus. Large
arrow indicates a cell with an invaginated nucleus;
small arrows point to two nucleoli in another cell.
The brain was embedded in plastic (methyl metha-
crylate), sectioned at 2 pm, and stained with thionine.

12

 

 

13

not measured by planimetry.

Measurements were taken from HPC cells in order that the cell areas
from the other nuclei might be expressed as a percentage of HPC cell areas
to control for shrinkage. Variation in HPC cell areas over time, however,
prevented the use of this manipulation (see Results).

Cell Sampling. Samples of 150 neurons per subnucleus were taken from

 

each animal. In PVl and PVm, 25 cells were sampled in each of six sections
chosen randomly from throughout each subnucleus. The same procedure was
followed for horizontal sections through SOa, SCdm, and SCvl. For coronal
sections through these subnuclei, however, 12 sections were chosen, and

12 or 13 cells were sampled per section. This manipulation insured adequate
representation through the extensive rostralcaudal limit of 80a, and com-
pensated for the high rejection rate in SCN, since the nuclear characteristics
of many of these cells were not discernable. Sampling was accomplished
using a Whipple-Hauser reticule in one eyepiece of a Zeiss microscope. A
particular nuclear group was placed under the grid, and before finally
focusing a decision was made as to which rows or columns would be sampled.
The objective was then focused and cells were counted. A 40x objective
with 12.5x eyepieces was used to sample cells from the PV1, PVm and 80a.
Samples from SCN were taken under oil immersion with a 100x objective and
12.5x eyepieces. For PV1, PVm and 50a, cells fell into one of the following
categories, having either a) one nucleolus, which was not in contact with
the nuclear membrane; b) one nucleolus, which was apposed to the nuclear
membrane (marginated); c) two or more nucleoli, both of which were un-
marginated ('multiples'); or d) two nucleoli, at least one of which was
marginated. In addition to these categories, cells of the SCdm and SCvl
were further divided into cells with or without a prominent nuclear invagin—

ation (i.e., was at least one third of the estimated nuclear diameter; see

14

Figure 2). Cells with invaginated nuclei were also compared to those
without in terms of the occurrence of marginated and multiple nucleoli.

For data analysis, morphological data were transformed to percentages.
The non-parametric Mann-Whitney test was used for between group comparisons
and the t-test for related means was used for within group comparisons. A
product-moment correlational statistic was employed to make further comparisons
between subnuclei, and between subnuclei and physiological data, within

groups.

RESULTS

Blood measurements

The data for plasma protein concentration and hematocrit are listed
in Table 1. Between group comparisons for protein concentration revealed
only one reliable difference: group 0600 had a significantly lower mean
than did groups 1200 and 2400. While there was no overlap in individual
hematocrit values between groups 0600 and 1200, the loss of one sample
prevented a meaningful statistical comparison. All other groups were
similar to one another on this measure.

Mean plasma corticosterone concentrations are plotted as a function
of time of day in Figure 3. A clear diurnal rhythm can be observed, with
a peak occurring at 1800 hr and a trough at 2400 and 0600 hr. Groups 2400
and 0600 had significantly lower corticosterone concentrations than did
group 1800 (p <.028, for each comparison), despite wide variation.within
groups. While exhibiting a periodicity, these values are higher than the
"unstressed" values reported by other investigators, presumably because
of the different blood sampling techniques employed (see Discussion).
Suprachiasmatic nucleus

As the photomicrograph in Figure 2 illustrates, cells of the SCN often
display multiple nucleoli and one or more marked, nuclear invaginations.
While in Nissl-stained preparations these infoldings appear as basophilic
stripes, electron microscopic examinations have confirmed the existence of
numerous nuclear indentations in SCN neurons (Lebeaux, 1971; Subaro and
Pellegrino de Iraldi, 1969). The presence of these membrane extensions,

the small size of the cells, and the highly granular appearance of the

15

16

Table 1. Plasma protein and hematocrit values at different times of
of the day (x :1; SEN)

Hematocrit_(Z Plasma)

Protein (g/100 ml)

Time

0600 60.3 i 2.36 5.4 i .31*
(lights on)

1200 51.3 i .63 6.5 i .06

1800 55.0 :- l.30 5.8 1 .21
(lights off)

2400 56.0 + 2.04 6.4 + .18

 

* Significantly less

than at 1200 or 2400 hrs, p < .028 for each comparison.

Figure 3.

Figure 4.

17

In the top panel the mean percentages of cells with multiple
nucleoli are graphed for the supra0ptic nucleus (SOa) as a
function of time of day. Mean corticosterone values (in ug/
100 m1 of plasma) from cardiac blood are similarly plotted
in the bottom panel. In this and in all other figures, the
dark bar indicates the period of darkness. Note the peak
values at 1800 hr, and the troughs at 2400 and 0600 hrs.

The mean percentages of cells exhibiting a nuclear invagina-
tion in the dorsomedial (SCdm) and ventrolateral (SCvl)
portions of the suprachiasmatic nucleus are plotted as a
function of time. SCvl exhibits significantly higher values
at all times, except at 0600, where convergence can be ob-
served.

18

 

 

 

 

A209 _.._Ow..oDz
MJQEHSE .x.

d 4

mo 0 0
5 I

3

258:2:
mzommemoofimoo

2400
I800
TIME OF DAY

IZOO

OGOO

 

 

o SCdm
o SCvI

 

80

mo mu m
7 6 5

_m.._oDz om._.<z_o<>z_ ox.

I
0
4.

2400
IBOO 0600
TIME OF DAY

IZOO

19

Figure 5. In the top panel the mean percentages of 'multiples' for all
sampled cells in either the dorsomedial (SCdm) or ventrolateral
(SCvl) portion of the suprachiasmatic nucleus are graphed as
a function of time of day. Depicted in the middle panel are
similar values for cells exhibiting and invaginated nucleus.
In the bottom panel the mean percentages for cells without a
nuclear invagination are also plotted. Cells with invaginated
nuclei exhibit some variation in their percentage of 'multiples,'
with a peak 1800. Cells without invaginated nuclei exhibit no
such variation.

Figure 6. Depicted are mean cell areas (in umz) in the dorsomedial (SCdm)
(top panel) and ventrolateral (SCvl) (bottom panel) portions
of the suprachiasmatic nucleus as a function of time of day,
represented by solid line (scale to the leffi). Also shown are
the mean values of difference scores (in um ) from each sub-
nucleus obtained by subtracting the mean cell area of each
plane from each animal's mean value, represented by the
broken line (scale to the right). Cell areas peak at 1800
hr for both subnuclei. The changes in cell area appear
much more profound when the differences between planes are
minimized, as can be seen by examining the difference scores.
H: cells taken from horizontal plane, C: cells taken from
coronal plane.

% MULTIPLE NUCLEOLI

CELL ARE A (usz

 

 

 

 

 

 

0 SCdm
4o- 0 SCVI
30- g
Total
40"
30‘
lnvog. nuc.
40~
30- >974:
Noninvog. nuc.
201%
I200 2400
IBOO 0600
TIME OF DAY
SCdm
80‘ 1 b*IO

 

 

 

I200 2400
IBOO 0600

TIME OF DAY

21

nucleoplasm accounted for a high rejection rate during cell sampling.

Invaginated nuclei. In addition to their locations within the SCN,

 

SCdm and SCvl were also distinguished by their respective populations of
cells with invaginated nuclei (see Figure 4). Significantly more of these
cell types were found in SCvl than in SCdm across all groups (t=4.39, d.g.=
15, p < .001), though in Figure 4, a convergence can be observed at 0600
hr. A rhythmic occurrence of these neurons is suggested by Figure 4, but
variability allowed only one difference to attain statistical significance.
In SCvl, the mean percentage of cells exhibiting invaginated nuclei was
higher at 1800 than at 0600 hr (p < .028).

Multiple nucleoli. As depicted in Figure 5, there were no dramatic
fluctuations in the percentage of 'multiples' in either subnucleus as a
function of time. Accordingly, no statistical differences were found in
this regard. However, when cells with invaginated nuclei were considered
separate from those without, some differences emerged. Cells with an infolded
nucleus seemed to vary in their percentage of 'multiples' in both SCdm.and
SCvl. Only SCvl exhibited a significant difference, with more of these cell
types present a 2400 hr when compared to 1200 hr (p < .028). This response
was not significantly correlated with corticosterone concentration in either
SCvl (1:!3 -.l3) or SCdm (r- .09). Neurons without invaginated nuclei, as
can be seen from Figure 5, did not vary significantly in their percentage
of 'multiples'. '

Marginated nucleoli. When no distance could be perceived between a
nucleolus and the nuclear membrane, the former was considered marginated.
Since a nuclear infolding represents a part of the nuclear membrane, nuc-
leoli in close apposition with these basophilic stripes were considered
marginated. Listed in Table 2 are the data for marginated nucleoli in

SCN. Populations were distinguished on the basis of invagination, number

22
of nucleoli, and location within the SCN. No significant changes were

found across times in the occurrence of margination, regardless of
category.

Table 2 also shows significant within group comparisons. When comparing
cells with and without invaginated nuclei, it was found that, in cells with
only one nucleolus, margination occurred more often in the former for
both SCdm (t= 5.33, d.f.- 15, p < .001) and SCvl (t= 7.912, d.f.- 15,

p < .001). Although not as great a differenCe was found, the same was
true for 'multiples' in SCdm (t= 2.876, d.f.- 15, p < .02) and SCvl (t-
2.959, d.f.= 15, p < .01). It thus appears that a nuclear infolding in-
creases the probability of nucleolar margination. Since SCvl contained
more cells with invaginated nucleoli, the hypothesis was tested that
margination was more prevalent in this subnucleus when compared with SCdm.
When all cells with marginated nucleoli were totalled, however, no difference
was obtained between the two portions of SCN. This rough equivalence was
apparently due to the fact that, in SCdm, more cells without invaginated
nuclei contained marginated nucleoli than in SCvl (t= 2.839, d.f.' 15,

p < .02). No other within group differences were noted.

.Esll”§i£2° Regression equations were obtained for both subdivisions,
and for both planes, to predict cell area. In SCdm, the weighted average
of the long and short axes was highly correlated with cell areas determined
by planimetry, yielding a coefficient of .908 in the coronal plane, and
one of .902 in the horizontal plane. From these relationships the equations
y - .2935x - 5.652 and y - .2654x - 3.738 were derived, respectively. A
similarly high correlation was obtained for SCvl. In the coronal plane,

r - .844 and the derived equation was y = .147x - 1.297; in the horizontal

plane, r - .848 and the derived equation was y - 2.17 - 2.924.

23

.Ho. v a .moaawuaoa
.No. v a .mOHQAuaoa
.Hoo. v a .moawaam
.aoo. v a .moawaam

.mpfia Sow
.w>ufia scum
.338 now
SEE soom

mono nouooum mauamoamuswwm
mono noumouw hauuuowwaawflm
umou Monmoum hfluduoamadwam
ammo “museum hauawoamwnwam

OO
O

«i
fl

msoaooa wouwcwwu>na no uaonuaa mHHoo I .w>nwm
mamaoom woumnfiwm>nfi cm spas mHHuo I .w>nw

 

 

 

 

 

 

 

 

 

 

moo Hooo ooooq Moos so “mom «.om. No.2. do Hoé 03453 om.~ HES «.24 HTS .um.
o2 H54... 8. HES was No.2 35 No.8 so... #4... 3.. No.3 35 Hmoo 26 No.2 8%
93 H3: is No.3 35 Moon o.m HAS $6 Hmoomoé Hogs So No.2 2.“ No.8 83
.34 Hwém $4 H~.3 «Tm Hméo mH.~ Hoéw «to. Hwéo «A Neda wmé Hméf owé Hmdm 83
$6 Hioo oo. Moos an; whom 24 Home $4 H33 3. Moos So HT: on. Hooo oooo
.w>cwn .w>s« m>ch .w>a« .m>caa _ .m>aw .w>cwa .w>a«
mmflmfiuanz mmawawm mmamwuanz .mmwwmflm
Son 58

.hmv onu mo mmawu unououmav um H>om
new Boom ma «Hooaoac woumsawuma soak wHHou no owmuaoouom Axum.Hv duo:

.N manwa

24

Scant amounts of cytOplasm were observed when SCN perikarya were
projected for tracing. In many cases it was not known whether the nuclear
or perikaryal membrane was drawn. There were no differences found between
groups in this difficulty, so that all samples represent similar popula-
tions of cells in this respect. It was not possible to present cell areas
in SCN as a percentage of HPC control cells, since the latter varied in
size as a function of time (see Figure 8).

Figure 6 shows the mean predicted cell areas for SCvl and SCdm
across time. Both portions of the SCN exhibit fluctuations in the size
of their constituent cells. Wide differences between areas in each plane
prevented a statistically significant result, however. Horizontally sec-
tioned cells were larger than coronally sectioned cells in SCdm and SCvl
(p < .0001,for both subnuclei). In an attempt to remove this difference
between planes so that between group comparisons could be made, the mean
cell area across all groups was calculated for each plane within each
subnucleus. This mean was then subtracted from each individual score.
This provided a quasi-standardization of scores, and the result of this
manipulation can be seen in Figure 6. Similar peaks and troughs occurred
in the magnitude of these differeme scores as in cell size, but the
changes were more dramatic. Difference scores from groups 1800 and 2400
were significantly larger than those at 1200 hr in both subnuclei (p < .028,
for each comparison). I

Cell area proved to be still another basis on which the neurons of
SCdm and SCvl could be distinguished. Larger somal areas were found in
SCvl in both horizontal (t- 5.334, d.f.= 7, p < .001) and coronal (t- 8.333,
d.f.- 7, p < .001) planes.

Since the vacillations in cell area in the SCN were somewhat in

phase with the corticosterone rhythm, a correlational measure was applied

25

to see if this temporal relationship was significant. For coronally
sectioned SCdm cells, r= .731 (d.f.- 6, p < .025). In horizontally sliced
brains, however, cell areas in SCdm.were not significantly correlated with
corticosterone concentrations (r- .441). A similar trend was observed in
SCvl, with a high positive correlation found between cell areas in the
coronal plane and plasma corticosterone (r= .787, d.f.- 6, p < .025) and

a non-significant one between the same two variables in horizontally
sectioned brains (r= .248).

Supraoptic nucleus

Magnocellular neurons of SON are different in appearance as well
as in size from the parvocellular elements of SCN. Their morphology in
Nissl-stained tissue is essentially like that of the PVN perikarya, shown
in the photomicrograph in Figure 7. Large amounts of deeply-stained cy-
toplasm.are evident. While the nucleoplasm is somewhat granular, nucleoli
stand out as dark spheres in the round or oval nucleus. Invaginations
are rare in the nuclear membrane.

Multiple nucleoli. As graphed in the top panel of Figure 3, the
percentage of 'multiples' in the SOa showed a distinct diurnal rhythm
paralleling that of corticosterone secretion, characterized by a positive
correlation coefficient (rn .70, d.f.- 14, p < .01). Nucleolar pro-
liferation peaked at 1800 hr, with a precipitous decline by 2400 hr, and
a subsequent trough at 0600 hr. ‘Mean values of 'multiples' were signifi-
cantly lower at 2400 and 0600 hr when compared with group 1800 (p < .028,
for each comparison).

Marginated nucleoli. As with the parvocellular SCN neurons, nucleoli
in 80a were often seen in close proximity with the nuclear membrane. Data
for marginated nucleoli in the magnocellular PVN and 80a are listed in

Table 3. In the SOa, no differences were found between groups in the per-

Figure 7.

26

Photomicraph of a 6 pm coronal section through the para-
ventricular nucleus, illustrating magnocellular neurosecre-
tory neurons, some of which contain multiple nucleoli (large
arrows). Small arrows point to marginated nucleoli. The
brain was embedded in paraffin and stained with thionine.

27

 

28

.Hoo. v a .moawsww wouoswmuma a>m Hmuou swan amumouw haunuowuwawam as

.wuo. v a .ua oowa um moaawuasa woumowwuma abm mono wouwouw hauamowmwawwm e

 

 

 

No No.3 too No.2 o.o Hfioo o.m No.2 o.o +35 o4 Hos... mm. 153
so Hog: mo HTS no Hfimo To HES No “moo do Moos. oosu
I I I I I I Coo REES
To +~.oo hm +o.mm m... +N.No ma 3.2 so + «so To +33 83
m.~ No.5 in Hmém «o4 H3; o.~ Hod To HAS 2.. Hmon 82
I I I I I I Ado 333.8
in +m.oo an +9.; «A +m.~o m4 +92 n.~ 3.3 o.~ +o.om oooo
Saudi 3&on SEES: moflwfim «on3.32 mfimfim 33
:5 a: mom
.hso any mo
moses snowshoes so zoo one zoo «no no «Hooauau
ooumaawuma sues naaou mo omwuaoouoa ASMm_Hu now: .m manna

29

Figure 8. Mean cell areas (in pmz) from the supra0ptic nucleus (SOa)

Figure 9.

(top panel) and hippocampus (HPC) (bottom panel) are plotted
against time of day, and represented by solid lines (scale
to the left). Also shown are the mean difference scores

(in um ) from each nucleus, obtained by subtracting the mean
cell area of each plane from.each animal's mean value, rep-
resented by broken lines (scale to the right). In the top
panel, note the separate peaks in area for cells in the
horizontal (2400 hr) and coronal (1800 hr) planes. Total
cell area, as judged by difference scores, peaks at 1800 hr.
Note magnitude of the difference scores at 1800 and 2400 hrs.
In the bottom panel, HPC cell areas peak at 1800 hr, and
quickly drop off at 2400 hr. These changes are similarly
reflected in the difference scores. H: cells taken from
the horizontal plane, C: cells taken from the coronal
plane.

The mean percentages of cells with multiple nucleoli are
plotted for the lateral (PV1) and medial (Pvm) divisions of
the paraventricular nucleus as a function of time of day.

The PVl shows a higher percentage of 'multiples' at all times

‘than does the PVm. Neither the PVl or the PVm vary in their

percentages of 'multiples' significantly over time.

CELL AREA 01sz

3O

 

 

 

 

 

 

 

K
x' ‘. so
290- ' ‘ o ..20
-+l0
240‘ _ 0
--IO
I90'
' .-20
I4OJ ,
«40
290'
"20
240- _ O
l90‘ “20
I4OJ- "40
IZOO 2400
I800 0600
TIME OF DAY
_ o PVI
3 40‘ 0 PVm
Lu
4
8
Z 30" \\/
Lu
.1
o—- 20‘ \—$\O
'3
é’
32 IO-

I200 2400

I800 0600
TIME OF DAY

31

centage of cells with a single, marginated nucleolus. Likewise, the
percentage of 'multiples' with at least one marginated nucleolus did not
vary across time.

ggllugigg, The regression equations used to predict cell areas in
SOa were as follows. In the coronal plane, y. .368x - .901, where the
correlation between areas determined by planimetry and the weighted average
of the long and short axes was given by the coefficient .953. In the hori-
zontal plane, this equation was y= .353x - .812, and the correlation co-
efficient was .936.

Surprisingly, regular fluctuations in cell area were not dramatic in
the SOa. As seen in Figure 8, cell size does vary as a function of time.
Discrepancies between planes again prevented a valid statistical comparison.
Also, Figure 6 illustrates the dissociation between the peak values of cell
areas in the two planes. Coronally sectioned brains exhibited a peak in
cell area at 1800 hr, or at lights off. In the horizontal plane, however,
this peak.was observed at 2400 hr. "Standardized" difference scores are
also graphed in Figure 8. Despite what appear to be profound differences,
no significant deviations between groups were found in mean value of these
difference scores. No significant differences were detected in the size
of cells between planes.

Paraventricular nucleus

Figure 7 contains a photomicrograph of the lateral division of PVN
illustrating the basic characteristics of magnocellular neurosecretory
neurons. The description fits that given for SOa.

MUltiple nucleoli. That the two subdivisions of PVN differ in more
than just their relative locations within the complex is in part demonstrated
by comparing their proportions of cells with multiple nucleoli. As pre-

viously shown (Hatton et al., 1976), the PVl was found to contain a higher

32
percentage of 'multiples' than PVm across all groups (t= 6.19, d.f.= 15,

p < .001). Between group differences were not detectable, as neither
subnucleus changed in its percentage of 'multiples' as a function of the
time of sacrifice (see Figure 9). There was a significant correlation
between the two portions on this measure (r= .759, d.f.- 14, p < .001).

Marginated nucleoli. A significantly greater percentage of cells in

 

the PVl had a single, marginated nucleolus when compared with Pvm (t- 4.85,
d.f.- 15, p < .001; see Table 3). Margination in cells with just one
visible nucleolus did not change significantly across times. In the Pvm,
however, cells with multiple, marginated, nucleoli were more prevalent at
1200 hr than at 1800 hr (p < .028). Cells within the PVl of this type
showed no such response. No significant differences were found between
the two subnuclei in their relative proportion of 'multiples.'

.ggllugigg. Figure 10 depicts the mean cell areas in the PVl and Pvm
over time, predicted from the following equations. For PV1, y= .299x - 5.577
in the coronal plane, where r- .895; in the horizontal plane, the coefficient
of correlation between cell areas and the weighted average of the long and
short axes was .949, where y- .395x — 10.205. For the PVm, r= .939 in the
coronal plane and y= .298x - 5.797; rII .945 and y= .332x - 7.233 in the
horizontal plane.

Diurnal oscillations in cell areas from both subnuclei were profound,
particularly in the horizontal plane where 100 um2 separates the cell areas
at 2400 and 1200 hr. The uniformity of this cellular response within the
PVN complex is the basis for a high correlation between the cell sizes
of two divisions (r- .92, d.f.- 15, p < .001). Combining planes, the mean
cell areas were greater at 1800 and 2400 hr than at 1200 hr in the Rvm and
PVl (p < .028, for each comparison). Examining within planes, more

differences seemed likely. Scores were thus standardized as described be-

33

fore for SCN and 80a to ferret out these changes, and are graphed in
Figure 10 as well. Difference scores showed a peak at 2400 in both por-
tions of the PVN, with a trough at 1200 hr. In both PVl and PVm, the
following differences were found significant. Groups 1800 and 2400 hr
had higher mean difference scores than groups 0600 and 1200 (p < .028,
for each compariton).

The rhythmic variation in cell size in the PVN is out of phase with
plasma corticosterone concentrations. Correlations were thus poor be-
tween corticosterone and cell areas in both the PVl (r- .44, p > .05)
and PVm (r3 .09, p > .05).

Further differences were found in cell areas between the PVl and
PVm. Lateral cells were larger than medials in both the coronal (t- 4.597,
d.f.- 7, p < .01) and horizontal (t= 6.5, d.f.- 7, p < .001) planes. Within
groups, cells from the horizontally sliced brains appeared larger than
those from the coronal plane, but no significant difference was detected.
H1pp9campal_pyramidal cells

Control cells taken from CA 4 of the hippocampus showed a marked
rhythm in their area (see Figure 8), preventing their use as controls for
cell shrinkage due to vagaries in embedding. Combining planes, variability
precluded any statistical differences from.being detected, so these scores
were standardized also (for predicted areas, the regression equations used
were y- .294x - 5.652, r- .908 in the coronal plane; for the horizontal
plane, y- .265x - 3.738, r- .902). The difference scores were significantly
higher at 1800 than at 1200 and 2400 (p < .028, for each comparison). The
rhythm for cell area changes was in phase with that of corticosterone
secretion, but individual values were not significantly correlated (r!3 .441,

p > .05).

Figure 10.

34

Mean cell areas (in umz) from the medial (PVm) (top panel)
and lateral (PV1) (bottom panel) divisions of the para-
ventricular nucleus are graphed as a function of time of
day, and represented by solid lines (scale to the left).
Also shown are mean difference scores obtained by sub-
tracting the mean cell area of each plane from each
animal's mean value, represented by broken lines (scale

to the right). The peaks in cell area in both the PVl
and the PVm occur at 2400 hr. This change is punctuated
further by a similar change in the mean difference scores.
H: cells taken from the horizontal plane, C: cells taken
from the coronal plane.

CELL AREA (11sz

 

35

 

 

 

IZOO 2400
IBOO 0600
TIME OF DAY

DISCUSSION

Concerning the results of this investigation, it is clear that a
diurnal cyclicity, based on the morphological indices examined, character-
izes the function of the hypothalamo-neurohypophysial system. Also, the
methodology, in particular the dependent variables one chooses to measure,
clearly influences what conclusions can be drawn. For instance, within the
80a and PVN, cell area and nucleolar proliferation were dissociated, and,
had either been the sole index of neurosecretory activity, different con-
clusions would have undoubtedly been reached. Furthermore, discrepancies
between this and previous studies of a similar nature may be attributed
to methodological differences, and will be discussed where applicable
below.

While not the major result of this investigation, it is useful to
consider the periodicity of plasma corticosterone first. Changes related
to neurosecretion may then be referred back to this important rhythm.

Plasma Corticosterone

 

Fluorometrically determined concentrations of plasma corticosterone
over the 24 hour period closely approximated the well established diurnal
rhythm, with a peak at lights off and a trough near lights on (Critchlow,
et al., 1963; Guillemin et al., 1959; Retienne, et al., 1968). Absolute
values at the peak and trough were noticeably higher than those reported
in these previous studies. At the peak, these elevations resembled levels
obtained from ether-stressed female rats at lights off (Zimmerman and
Critchlow, 1967). Comparable increments in corticosterone secretion were

detected by Zimmerman and Critchlow (1967) when rats were stressed at the

37
peak or trough of the rhythm, whereas here trough levels were only

slightly above normal. According to the results of Dunn, et al., (1972),
greater stress-induced increases should occur during the trough of the
corticosterone rhythm. Divergent results such as these make interpreta-
tions of the present data difficult, especially since they were obtained
from female rats, whose basal plasma corticosterone concentrations, and

' elevations to stress, are higher than those seen in males (Critchlow,

et al., 1963). Blood samples must be obtained quickly under ether
anesthesia (< 3 minutes) to avoid stressed corticosterone levels (Davidson
et al., 1968), and it is probable that all rats in the present study were
not sampled under the same intensity of stress.

Supraoptic and Paraventricular nuclei

 

Comparisons with previous studies. Increased cell areas have pre-
viously been reported in the rat SOa (Armstrong et al., 1977; Hatton and
Walters, 1973) and PVN (Hoblitzell, et al., 1976) under conditions known
to induce hormone production. Although choosing only ten of the "largest"
cells per animal in an unspecified plane of section, Glantz (1967) found
SON cell areas larger at lights off that at lights on. A similar trend
is evident in Figure 6, but predicted somal areas varied in their peaks
depending on the plane of section. In both the PVl and PVm, the observed
peaks in cell area (see Figure 10) match those for estimated nuclear volume
reported by Rinne and Sonninen (1964).

Whereas Glantz (1967) was unable to detect fluctuations in the density
of staining in Gomori-positive SOa cells in the rat, Rinne and Sonninen
(1964) subjectively observed diurnal differences in the density of staining
in SON and PVN neurons using aldehyde fuchsia. The SON and PVN were both
stained lightly between 2400 and 0400 hrs, a time when nuclear volumes

were reportedly largest in both nuclei. Unfortunately, their study was

38
run when little if any darkness was available. This deficiency detracts

from their conclusion that the nuclei were more active at night. Other
evidence indicates that plasma levels of VP are higher in the morning

(four hours after lights on) than at night (Mbhring and MBhring, 1975).

It has also been found that albino rats excrete small amounts of con-
centrated urine during the daytime, in contrast to the larger, more dilute
quantities collected at night (Glantz, 1967). Densitometric measurements
of neurosecretory material are poor as an index of neurosecretory activity,
because they might vary with cell size, regardless of changes in dense
core vesicle content, and could thus veil the true dispersion of neuro-
secretory material within the perikarya. It is possible that such a
methodological inadequacy made major contributions to the discrepant findings
of Rinne and Sonninen (1964).

Megphological indices and neurosecretion. Proliferation of nucleoli
occurs in rat SOa neurons subsequent to osmotic stimulation, either in
combination with plasma volume depletion, i.e., during water deprivation
(Hatton and walters, 1973), or alone, as with hypertonic saline ingestion
(Armstrong, et al., 1977). The peak level of 'multiples' during the diurnal
cycle indicates the SOa has been most active during the day. This response
could be less related to secretion per se, than to the restoration of VP
(or OX?) in the depleted neural lobe. The fact that pituitary concentrations
of VP are lowest during the morning supports the notion that nucleolar
multiplication is out of phase with VP secretion by at least six hours.
However, the lack of information regarding a diurnal rhythm in 0X secretion
and a complete description of VP levels in the blood over the day, make a
complete assessment of this response impossible. If PVN neurons participate
in diurnal VP secretion, then cell areas in this nucleus are also out of

phase with circulating and pituitary VP titers. Given the rostro-caudal

39
distributions of 0X and VP within the PVN (Swaab, et al., 1975), the

homogeneity of the areal response across both subnuclei implies OX and VP
cells were active at the same time.

Origin gf_rhythmicity. There are few clues as to the origin and

 

control of the daily fluctuations in magnocellular neurosecretory function.
Neither the PVN or SON are thought to receive retinal input, and Swanson
and Cowan (1975) failed to observe any SCN projections to the two nuclei.
According to Nobin, et al., (1972), dopaminergic fibers from the rat
periventricular and arcuate regions, both of which receive SCN input (Swanson
and Cowan, 1975), innervate the neurohypophysis. Dopamine elicits a graded
release of OX and VP from the rat hypothalamus, including the pituitary
stalk, $3 vi££2_(Bridges, et al., 1976). It is tempting to speculate that
neurohypophysial hormone release may be effected by direct inputs on to
neurosecretory axons or terminals, and that this release may thus oscillate
over the day. The observed changes in nucleoli and cell size may reflect
hormone production that is accomodating to neural lobe depletion. Such an
hypothesis would account for phase differences between the cell responses in
this study and VP levels in pituitary and plasma over the 24 hour period.

Increases in plasma protein concentration accompany blood volume de-
pletion, which is an adequate stimulus for ADH release (Fabian, et al.,
1969). Even though changes in plasma protein as measured by refractometry
were evident in the present study, it cannot be determined whether or not
these reflect blood volume depletion. The data for hematocrit, which can
also be used as an index of blood volume, were incomplete where differences
were suggested. It is possible that such an internal signal participates
inIor results from, the activation of the SOa or PVN during the diurnal
cycle.

Relation £g_corticosterone. From the cycles in 80a and PVN morphology

4O

accounted for in this study, a role in adenohypophysial function is comr
patible for either nucleus. There are changes in large and small dense

core vesicles in the SOa after four hours of water deprivation, when both
small (< 1600 X) and larger (> 1600 2) vesicles are depleted in the cytoplasm
(Tweedle and Hatton, 1976). Electron microscopic investigations have confirmed
the existence of neurophysin— and VP-positive granules in the fibers of the
MEx which are mostly in the 900-1100 X range (Silverman and Zimmerman, 1975).
Following adrenalectomy in the rat, Dube, et al., (1976) noted an increase

in the number of small VP granules within axons abutting portal capillaries.
Whether or not these small vesicles originate within the magnocellular
neurosecretory perikarya, or perhaps the SCN, remains to be determined. A
role in CRF secretion for either the PVN or SOa would again imply that
fluctuations in hormone production are out of phase with actual secretion,
since increases in CRF secretion would appear prior to those in corticos-
terone.

Differences between SON and PVN. Demonstrations of OX and VP in

 

near equal proportions have led recent authors to postulate a functional
similarity between the two nuclei (George et al., 1976). Such an hypothesis
seems untenable with the absence of nucleolar proliferation in the PVN in
light of its large cell size changes, while the SOa exhibits much smaller
deviation in mean cell area, yet shows a distinct rhythm in its percentage
of hultiples.’ Perhaps areal and nucleolar changes are simply the equivalent
adjustments cells in each nucleus make, respectively, to produce hormone.
This is unlikely since both nuclei respond with increases in somal area

and in the percentage of 'multiples' during water deprivation (Hatton and
Walters, 1973; Hoblitzell, et al., 1976). Given the elevated levels of
corticosterone due to stress, the nucleolar response in the SOa may have

been imposed by ether inhalation and cardiac blood sampling. Animals given

41

subcutaneous injections of isotonic saline when restrained show a short
latency increase in the percentage of cells with multiple nucleoli in
the SOa (Armstrong, et. al., 1977). However, the possibility that the
percentage of 'multiples' in the SOa was stress-induced would still re-
present a difference between the two nuclei.

Recent anatomical inquiries have demonstrated separate afferent and
efferent connections of the two nuclei. In particular, the PVN issues
axons to the caudal brain stem and spinal cord, as well as to the neuro-
hypophysis in the rat (Saper, et al., 1976). In the monkey, unilateral
lesions of the PVN deplete VP from the ipsilateral MEx (Antunes, et al.,
1976). The changes in cell area in the PVN may then reflect extra-
neurohypophysial activity, especially that which may be related to
adenohypophysial ACTH release.

Suprachiasmatic nucleus

Diurnal changes. No profound changes in any of the variables
measured in SCN were witnessed. Between group differences that were
evident were small and are Open to varying interpretations. Differences
between SCdm, whose cells contain VP and its associated neurophysin, and
SCvl, were restricted to within subjects, and did not provide clear
evidence for a functional separation of the two divisions. Criteria used
in this study for neurosecretory activity, derived from responses in the
magnocellular neurosecretory system, may not have been sensitive to the
secretory capacity of SCdm cells.

Cells with invaginated nuclei do show some variation in their per-
centage'of 'multiples' over time, particularly in the SCvl. These cell
types account for approximately half of the sampled cell pOpulations from
both subnuclei, but are themselves labile over time. Changes within SCvl

on this measure, and in cell size, are probably to related neural activity

42

in general, as SCN is involved in the maintenance of many bio-behavioral
rhythms (Moore and Eichler, 1976). If real, the apparent changes in cell
area in the SCdm may reflect neurosecretory activity which could account
for CRF secretion.

SCvl and SCdm. The differences between SCdm and SCvl on several

 

measures are interesting butuninterpretable at this time. In the rat,
retinal ganglion cell fibers enter the ventromedial protion of the caudal
aspect of SCN, and the possibility is slim that these axons terminate on
VP or neurophysin producing cells (Mason and Lincoln, 1976). Extensive
intrinsic connections have been described within the rat SCN (Guldner,
1976), so a close but indirect photoperiodic modulation of hormone-con-
taining cell bodies in the SCdm may exist. No information is available
which might indicate a tOpographic distribution of terminals from raphe'
nuclei within the SCN. Detailed anatomical and physiological investigation,
especially with the use of the electron microsCOpe, are needed to determine
more precisely the differences between the SCvl and SCdm under different
conditions. The synthetic apparatus and product in the cytoplasm of SCN
cells may be sensitive to slight increases in secretion. For example,
Clattenburg, et al., (1975) have found changes in the Golgi complex, as
well as in the dense core vesicle population, in the rabbit SCN after
ovariectomy. To what hormone these changes are related is as yet un-
known.

_R_o_l_g _o_f_ VP; in _S_(_3_N_. Due to a lack in species generality, the role
of the VP in the rat SCN in ACTH regulation has been questioned (Silverman
and Zimmerman, 1975). Only the rat and mouse SCN have been found to contain
VP or neurophysin (Zimmerman, 1976). As discussed above, the SCN’may
mediate diurnal neurosecretory rhythms indirectly through its connections

with the arcuate, ventromedial, and periventricular regions of the hypothalamus.

43
It is interesting that Slusher (1964) selectively abolished the corticos-

terone rhythm in the rat with lesions in the periventricular area, leaving
adrenal responses to stress intact. Lesions made more basally only pre-
vented the stress response. The SCN input to the periventricular area
may be necessary for the nucleus to affect CRF secretion. Perhaps VP
serves as a neurotransmitter in this regard.

Hippocampus

The phase relationship of plasma corticosterone with the area of
HPC cells is probably more than coincidental. 3H—corticosterone is taken
up in large quantities by the cell nuclei in hippocampus (Whrembourg, 1975).
The physiological effect of this hormone on these cells has not been fully
determined, but in hippocampus some role in modifying unit activity under
different behavioral conditions is likely. For example, theta cells (cells
intimately associated with theta potentials) increase their firing rates
during certain behaviors when awake as well as during paradoxical sleep
(Ranck, 1973). In the case of nocturnal animals, corticosterone may
participate in preparing the organism for activity, with its peak near
lights off.

Although disruption of the corticosterone rhythm has been reported
following sections of the fornix by Moberg et al., 1971, others (Wilson
and Critchlow, 1974) have failed to replicate this finding. While lesions
of the hippocampus have no effect on corticosterone rhythmicity (Lanier
et al., 1975), stimulation has been reported to inhibit ATCH release under
basal conditions (Mangili et al., 1966). The evidence is as yet in-
conclusive, but these studies indicate some involvement of the hippocampus
in pituitary—adrenal function. The heavy binding of corticosterone to
hippocampal cell nuclei suggests this structure may participate in a long

feedback loop, with its output affecting hypothalamic function, and perhaps

44
even modulating CRF release. PrOper electrophysiological studies,

however, need to be carried out concurrently with hormone assays in
order to more precisely determine the effect of corticosterone on hippo-
campal cell function, and the effect of the latter on CRF secretion.
3222.11

In the magnocellular neurosecretory nuclei, demands are met resulting
from increased hormone release by adjustments in organelles responsible
for RNA and protein synthesis. The nucleolus serves to produce much of
the high molecular weight ribosomal RNA vital to protein production (Busch
and Smetana, 1970; p. 3). Nucleolar proliferation in SOa (Armstrong et al.,
1977; Hatton and welters, 1973) can be induced by osmotic stress, known
to initiate enhanced 3Hr-leucine incorporation (Norstrom, et al., 1971) and
an increase in the uptake of labelled cytidine (George, 1973) in the nucleus.
This nucleolar response thus appears correlated with indices of protein or
RNA synthesis. The formation of additional nucleoli is an adaptive response,
but in mature neurons the mechanism of this multiplication is unknown. A
nucleolar chromosome contains a region designated the nucleolus organizer,
which houses the DNA templauaresponsible for RNA synthesis as well as
nucleolar formation (Busch and Smetana, 1970; pg. 116). No outstanding
ultrastructural differences can be seen between two nucleoli within a single
nucleus in SOa neurons (Tweedle, personal communication). Studies are
now in progress in our laboratory utilizing the uptake of 3H-uridine during
stimulated states as a measure of activity. More labelled material in
'multiples' when compared with cells having a single nucleolus may indicate
an enhanCed protein synthesizing capacity of the former.

In previous investigations of nucleolar proliferation in our lab-
oratory, it has been noted that the increased percentage of 'multiples'

in the rat SOa induced by water deprivation (Hatton and Walters, 1973),

45

stress and hypertonic saline ingestion (Armstrong et al., 1977), remains
elevated for some time following termination of the activating stimuli.
For example, after 10 days of rehydration following 24 hours of water
deprivation, the level of 'multiples' fails to reach that observed in
nondeprived control rats (Hatton and Welters, 1973).

In the present study, the percentage of cells with multiple nucleoli
in the SOa shows dramatic shifts within six hours. Since in this case
the excitatory stimuli are as yet unknown, their onset and termination
cannot be determined. Nevertheless, it is interesting that these fluctua-
tions, whereby nucleoli must be removed and produced in relatively large
numbers, occur in the normal course of a day. Perhaps the neurosecretory
system is "conditioned" to such daily alterations, anticipating the level
of synthesis necessary to restore hormone levels in the neural lobe. When
under severe challenge, as with water deprivation, the hormone content
of the neurohypophysis is undoubtedly depleted beyond the restorative
capacity of the 80a (or PVN); all the produced hormone is being secreted.
Such a depletion would thus require prolonged synthesis until neurohypophysial
levels are returned to baseline levels.

The fate of produced nucleoli is unknown. Expulsion of nucleoli
through the nuclear membrane has been noted in nerve cells, but may be a
histological artifact (Busch and Smetana, 1970; pg. 429). Structures re-
sembling nucleoli in shape, such as nematosomes, or nucleolus-resembling
bodies, have been observed with the electron microscope in the cytoplasm
of 80a neurons (see LeBeaux, 1971; Tweedle and Hatton, 1975). The re-
lationship of such elements with nucleoli is poorly understood. The number
of such inclusions apparently does not increase during rehydration (Tweedle
and Hatton, 1975). Le Beaux (1971) speculates that nucleolus-resembling

bodies in the cytOplasm of neurosecretory cells are related to protein

46
synthesis in their own right, and are sufficiently distinct in their

ultrastructure from nucleoli to be considered as separate entities. As

can be gleaned from the preceding discussion, the formation and disposal
of new nucleoli in mature neurons are phenomena only recently documented;
much research is needed to clarify the mechanisms involved in these pro-

(288888.

LI ST OF REFERENCES

LIST OF REFERENCES

Aghajanian, 6., Bloom, F.E., Sheard, M.H.: Electron microscopy of
degeneration within the serotonin pathway of rat brain. Brain

Antunes, J.L., Carmel, P.W., Ferin, M., Zimmerman, E.A.: Paraventricular
nucleus: the origin of vasopressin secreting terminals on hypophysial
portal vessels in the monkey. Endocrinology,98, A324, (1976).

Arimura, As, Schally, A.V., Bowers, C.Y.: Corticotropin releasing activity
of lysine vasopressin analogues. Endocrinology, 84, 579-583, (1969).

 

Armstrong, W.E., Gregory, W.A., Hatton, G.I.: Nucleolar proliferation
and cell size changes in rat supra0ptic neurons following osmotic
and volemic challenges. Brain Res. Bull., in press, (1977).

Blackwell, R.E., Guillemin, R.: Hypothalamic control of adenohypophysial
secretions. Ann. Rev. Physiol., 35, 357-390, (1973).

Bridges, T.E., Hillhouse, E.W., Jones, M.T.: The effect of dopamine on
neurohypophysial hormone release in_vivo and from the rat neural lobe
and hypothalamus in vitro. J. Physiol., 260, 647-666, (1976).

 

Burford, G.D., Dyball, R.E.J., Moss, R.L., Pickering, B.T.: Synthesis
of both neurohypqnhysial hormones in both the paraventricular and
supra0ptic nuclei of the rat. J. Anat. (Lond.), 117, 261-269, (1974).

Burlet, A., Marche ti, J.: La vasopressine immunoreactive du noyou supra-
chiasmatique. Observations preliminaries 1e rat. C.R. Soc. Biol. (Paris),
169, 148-151, (1975).

Busch, H., Smetana, K.: The Nucleolus. Academic Press, New York, (1970).

 

Clattenburg, R.E.,‘Montemurro, D.G., Bruni, J.E.: Neurosecretory activity
within suprachiasmatic neurons of female rabbit following castration.
Neuroendocrinology, 11, 211-224, (1975).

Critchlow, V.: The role of light in the neuroendocrine system. In:
Advances in Neuroendocrinology (A.V. Nalbandov, ed.) University of
Illinois Press, Urbana, p. 377-402, (1963).

Critchlow, V., Liebelt, R., Bar-Sela, ML, Mountcastle, W;, Lipscomb, H.S.:
Sex differences in resting pituitary—adrenal function in the rat.
Amer. J. Physiol., 205, 807-815, (1963).

Davidson, J.M., Jones, L.E., Levine, 8.: Feedback regulation of adreno-
corticotropin secretion in "basal" and "stress" conditions: acute and
chronic effects of intrahypothalamic corticoid implantation. Endocrinology,
g2, 655-663. (1968).

 

47

48

Dierickx, R., Vandesande, F., May, J. de.: Identification, in the external
region of the rat median eminence, of separate neurophysin-vasopressin
and neurophysin-oxytocin containing nerve fibers. Cell Tiss. Res., 168,
141-151, (1976).

Dube, D., Leclerc, R., Pelletier, G.: Electron microscopic immunohistochemi-
cal localization of vasopressin and neur0physin in the median eminence
of normal and adrenalectomized rats. Amer. J. Anat., 147, 103-108, (1976).

Dunn, J., Scheving, L., Millet, P.: Circadian variation in stress evoked
increases in plasma corticosterone. Amer. J. Physiol., 223, 402-406, (1972).

Fabian,'M., Forsling, M.L., Jones, J.J., Lee, J.J.: The release, clear-
ance and plasma protein binding of oxytocin in the anaesthetized
rat. J. Endocr., 22, 175-189, (1969).

FUxe, R.: Distribution of monoamine nerve terminals in the central
nervous system. Actagphysiol. Scand., 22, Suppl., 247-37-85, (1965).

George, .M.: Localization in hypothalamus of increased incorporation
of H-cytidine into RNA in response to oral hypertonic saline.
Endocrinology, 22, 1550-1555, (1973).

George, J.M., Jacobowitz, D.M.: Localization of vasopressin in discrete
areas of the rat hypothalamus. Brain Res., 22, 363-369, (1975).

 

George, J.M., Staples, 8., Marks, B.M.: Oxytocin content of micro-
dissected areas of rat hypothalamus. Endocrinology, 22, 1430-1433,
(1976).

Glantz, R.: Circadian rhythms in the albino rat: Effects of illumination
on urine excretion, water intake, and the formation of the anti-diuretic
hormone. Physiol. Behav., 2, 49-54, (1967).

Guillemin, R., Dear, W.E., Liebelt, R.A.: Nychthermal variations in
plasma free corticosteroid levels in the rat. Proc. Soc. Exp. Biol.
Med., 101, 394-395, (1959).

Gflldner, F.H.: Synaptology of the rat suprachiasmatic nucleus. Cell.
Tiss. Res., 165, 509-544, (1976).

Hatton, G.I., Hutton, U.E., Hoblitzell, E.R., Armstrong, W.E.: MCrphological
evidence for two populations of magnocellular elements in the rat
paraventricular nucleus. Brain Res., 108, 187-193, (1976).

Hatton, G.I., Walters, J.E.: Induced multiple nucleoli, nucleolar mar-
gination, and cell size changes in supra0ptic neurons during de-
hydration and rehydration in the rat. Brain Res., 22, 137-154, (1973).

Hedge, G.A., De Wied, D.: Corticotropin and vasopressin secretion after
hypothalamic implantation of atropine. Endocrinolggy,122, 1257-1259,
(1971).

Hiroshige, T., Sakakura, M.: Circadian rhythm.of corticotropin releasing
activity in the hypothalamus of normal and adrenalectomized rats.
Neuroendocrinology, 1, 25-36, (1971).

49

Hiroshige, T., Sakakura,‘M., Itoh, S.: Diurnal variation of corticotropin
releasing activation in the rat hypothalamus. Endocrinoli,Japan, 22,
465-469, (1969). '

Hoblitzell, E.R., Hatton, G.I., Armstrong, W.E.: Paraventricular
nucleus: changes in the medial and lateral cell groups during
rehydration in the rat. Brain Res. Bull., 2, 329-332, (1976).

Hutton, U.E.: Changes in the paraventricular nucleus during different
stages of pregnancy and lactation in the albino rat. Unpublished
M.A..thesis, Michigan State University, (1975).

Koritsanszky, S., Koves, R.: Data on absence of axon terminals of medial
preoptic neurons in surface zone of median eminence. J. Neur. Transm.
22, 159-167, (1976).

Lang, E.F., Voight, R.E., Fehm, H.L., Pfeiffer, E.F.: Localization of
corticotropin-releasing activity in the rat hypothalamus. Neuro-
science Letters, 2, 19-22, (1976).

Lanier, L., Hartsveldt, C., Weis, B., Isaacson, R.: Effects of differential
hippocampal damage upon rhythmic and stress induced corticosterone
secretion in the rat. Neuroendocrinology, 22, 154-160, (1975).

Le Beux, Y.J.: An ultrastructural study of the neurosecretory cells of
the medial vascular prechiasmatic gland, the preoptic recess and the
anterior part of the suprachiasmatic area. 1.2. Zellforsch., 114,
404-440, (1971).

Lorente de No;, R.: Studies on the structure of the cerebral cortex. II.
J. Psychol. Neurol., 22, 113-177, (1934).

Mangili, G., Matta, M5, Martini, L.: Control of adrenocorticotrophic
hormone secretion. In: Neuroendocrinology, l. (L. Martini and
W.F. Ganong, eds.), Academic Press, New York. p. 297-370, (1966).

 

Mason, C.A., Lincoln, D.W.: Visualization of the retinohypothalamic pro-
jection in the rat by cobalt precipitation. Cell Tiss. Res., 168,

McCann, S., Dhariwal, A.: Hypothalamic releasing factors. In: Neuro-
gndocrinolggy_(L. Martini and W. Ganong, eds.), 1., Academic Press,
New York, p. 261-296, (1966).

Meirer, R.: Autoradiographic evidence for a direct retino-hypothalamic
projection in the avian brain. Brain Res., 22, 417-421, (1973).

Migeon, C.J., Tyler, F.E., Mahoney, J.P., Florentin, A.A., Castle, H.,
Bliss, E.L., Samuels, L.T.: The diurnal variation of plasma levels
and urinary excretion of l7-hydroxycorticosteroids in normal subjects,
night workers, and blind subjects. J. Clin. Endocr. Met., 16, 622-633,

1956).

Moberg, G., Scapagnini,U., de Groot, J., Ganong, W.: Effect of sectioning
the fornix on diurnal fluctuation in plasma corticosterone levels in
the rat. Neuroendocrinolggy, 7, 11-15, (1971).

50

MBhring, E., Mbhring, J.: Plasma ADH in normal Long-Evans rats and in Long-
Evans rats heterozygous and homozygous for hypothalamic diabetes in-
sipidus. Life Sci., 21, 1307-1314, (1975).

Moore, R.Y.: Retinohypothalamic projection in mammals: a comparative
study. Brain Res., 49, 403-409, (1973).

Moore, R.Y., Eichler, V.B.: Loss of a circadian adrenal corticosterone
rhythm following suprachiasmatic lesions in the rat. Brain Res.,

Moore, R.Y., Eichler, V.B.: Central neural mechanisms in diurnal rhythm
regulation and neuroendocrine responses to light. Psychoneuroendocrin-
ology, ;, 265-279, (1976).

 

Moore, R.Y., Lenn, N.: A retinohypothalamic projection in the rat.
J. Comp. Neurol., 146, 1-14, (1972).

Nashino, H., Koizumi, R., McC.Brooks, G.: The role of suprachiasmatic
nuclei of the hypothalamus in the production of circadian rhythm.
Brain Res., 112, 45-59, (1976).

Nobin, A., Bjorklund, A., Stenevi, 0.: Origin of the hypophysial catecholamine
innervation. Proc. Int. Cong,gEndocr., Wash. Excerpts Medics, No. 256.
New York, (1972).

Norstrfim, A., Enester, S., Hamberger, A.: Amino acid incorporation into
proteins of the supra0ptic nucleus of the rat after osmotic stress.
Brain Res., 22, 95-103, (1971).

Parry, H.V., Livett, B.G.: A new hypothalamic pathway to the median
eminence containing neurophysin and its hypertrophy in sheep with
natural scrapie. Nature, 242, 63-65, (1973).

 

Pearlmutter, A.F., Rapinto, E., Saffran, M.: The ACTH-releasing hormone
of the hypothalamus requires a co-factor. Endocrinology, 21, 1336-
1339, (1975).

 

Ranck, J.: Studies on single neurons in dorsal hippocampal formation and
septum.in unrestrained rats. I. Exp, Neurol., 22, 461-531, (1973).

Rinne, U.K., Sonninen, V.: Diurnal changes in the hypothalamoneurohypophysial
neurosecretion of the rat and its relation to the release of cortico-
trophin. Acta anat.,.2§, 131-145, (1964).

Retienne, R., Zimmerman, E., Schindler, W:J., Neuensehwander, J., Lipscomb,
H.S.: A correlative study of endocrine rhythms in rats. Acta Endocrinol.,
21, 615-622, (1968).

 

Saper, C.B., Loewy, A.D., Swanson, L;W., Cowan, W2M.: Direct hypothalamo-
autonomic connections. Brain Res., 117, 305-312, (1976).

Schally, A.V., Anderson, R.N., Lipscomb, H.S., Long, J4M., Guillemin, R.:
Evidence for the existence of two corticotrophin-releasing factors,
a and 8. Nature (Lond.), 188, 1192-1193, (1960).

51

Silverman, A.J., Zimmerman, E.A.: Ultrastructural immunocytochemical
localization of neurophysin and vasopressin in the median eminence
and posterior pituitary of the guinea pig. Cell Tiss. Res., 222,
291-301, (1975).

Singley, J., Chavin, W.: Cortisol levels of normal goldfish and response
to osmotic change. Amer. 2001., 22, 653, (1971).

Slusher, M.: Effects of chronic lesions on diurnal. and stress cortico-
steriod levels. Amer. J. Physiol., 206, 1161-1164, (1964).

Subaro, AAM., Pellegrino de Iraldi, A.: An ultrastructural study of
the rat's suprachiasmatic nucleus. J. Anat., (Lond.), 105, 439-446,
(1969).

Swaab, D.F., Pool, C.W., Nijveldt, F.: Immunofluoresceme of vasopressin
and oxytocin in the rat hypothalamo-neurohypophysial system. J. Neur.

Takebe, R., Sakakura, M., Mashimo, R.: Continuation of diurnal rhythmi-
city of CRF activity in hypophysectomized rats. Endocrinology, 22,
1515-1520, (1972).

Tigges, J., O'Steen, W.K.: Termination of retinofugal fibers in squirrel
monkey: a reinvestigation using autoradiographic methods. Brain Res.,

Tweedle, C.D., Hatton, G.I.: Ultrastructural comparisons of neurons of
supra0ptic and circularis nuclei in normal and dehydrated rats. Brain
Res. Bull., 2, 103-121, (1975).

Tweedle, C.D., Hatton, G.I.: The time course and reversibility of ultra-
structural changes in the nucleus circularis in dehydrating rats.
Neurosci. Abstracts II., 988, (1976).

Valtin, H.: Hereditary hypothalmic diabetes insipidus in rats (Brattleboro
strain). Amer. J. Med., 52, 814-827, (1967).

Vandesande, F., Mey, J. de, Dierickx, R.: Identification of neurophysin
producing cells 1. Cell Tiss. Res., 151, 187-200, (1974).

Vandesande, F., Dierickx, R., Mey J. de.: Identification of vasopressin-
neurophysin producing neurons of the rat suprachiasmatic nuclei. Cell
Tiss. Res., 156, 377-380, (1975).

Warembourg, M.: Radioautographic study of the rat brain after injection of
[1, 2-3H] corticosterone. Brain Res., 22, 61-70, (1975).

Wiley, K.M., Pearlmutter, A.F., Miller, R.E.: Decreased adrenal sensitivity
to ACTH in the vasopressin-deficient (Brattleboro) rat. Neuroendocrin-
ology, g5, 257-270, (1974).

Wilson, MTM., Critchlow, V.: Effect of fornix transection or hippocampectomy
on rhythmic pituitary-adrenal function in the rat. Neuroendocrinology,
22, 29-40, (1974).

52

Yasuda, N., Greer, M.A.: Studies on the Corticotrophin-releasing activity
of vasopressin, using ACTH secretion by cultured rat adenohypophysial
cells. Endocrinology, 22, 936-942, (1976).

Yates, F., Maran, J.: Stimulation and inhibition of adrenocorticotropin
release. In: Handbook of Physiology, IV. American Physiological
Society, Wash. D.G., p. 367-404, (1974).

Zimmerman, E.A.: Localization of hypothalamic hormones by immunocytochemical
techniques. In: Frontiers in Neuroendocrinology, 4. (L. Martini
and W. Ganong, eds.), Raven Press, New York., p. 25-62, (1976).

Zimmerman, E., Carmel, P., Husain, M., Ferin,‘M., Tannenbaum, M3, Frantz, A.,
Robinson, A.: Vasopressin and neurophysin: high concentrations in
monkey hypophysial protal blood. Science, 182, 925-927, (1973).

APPENDIX.A:

Materials

53
Materials

Blood Sampling;and analysis

1. Refractometer, American Optical, $125.00.

2. Centrifuge, International Equipment Company, $270.00.
3. Capillary Tubes.

4. Micro-capillary Centrifuge, Model MB, International Equipment Company,
$303.00.

5. Test tubes, VWR, $4.25.

6. Zoe glass syringes with 26 g needles.

7. 1cc tuberculin syringes, Plastipak.

Histology

1. Rotary Microtome, with knife; Arthur H. Thomas Company, $1600.00.

2. Thionine, Scientific Products, $40.00.

3. Paraffin, Paraplast, $2.10/2 1/2 lb.

4. 2" x 3" Microscope slides, Red label, $9.75/gross. Arthur Thomas Company.
5. Cover glasses, Red label, 1 oz. box, $4.10, Arthur Thomas Company.

6. Jars.

7. Base molds (1" x 1 1/2" x 1/2") $2.00/ea.; and embedding rings; $41.30/100 c.
M_i_¢_:roscope ecmipment

1. Zeiss microscope, Morgan Instruments, $1200.00.

2. Whipple-Hauser reticule.

3. Oil imersion lens, 1002:, $900.00, Morgan Instruments.

Photographic Eguipment

1. Camera: Dierdorf, S" x 7" format and optical bench arrangement.

2. Fibm: Kodak Contrast, Process Ortho.

APPENDIX B:

Histology

54
Histology

Paraffin embedding

l. Perfuse rat with saline (.90% NaCl) followed by a solution of 10%
formalin in .90% saline. Remove brain from skull and soak in the fixative
for several days.

'2. Place brain in a jar with one layer of cheese cloth covering the mouth,
secured by a rubber band. Submerge jar in container (tub or wash basin)
with running water overnight.

3. Block the brain to remove undesirable areas.

4. Begin dehydration, following the schedule listed below:

30% alcohol overnight
50% alcohol 2 hrs.
70% alcohol 2 hrs.
95% alcohol 2 hrs.
100% alcohol 1 hr.
fresh 100% alcohol 1 hr.
Methyl Benzoate
+ 100% alcohol
(equal volumes) 1 hr.
used Methyl Benzoate 1 hr.
fresh Mbthyl Benzoate 1 hr.
In oven at 60°C
fresh paraffin 1 hr.
fresh paraffin 1 hr.
fresh paraffin 1 hr.
fresh paraffin embed

5. To embed, poor a small amount of paraffin in a greased (vaseline or
silicone) metal base mold, and allow it to partially, but not completely,
solidfy.

6. Place brain on above layer of paraffin in base mold. Place embedding
ring on top of base mold, and fill mold with paraffin.

7. Allow paraffin to harden (an hour at least).

8. Remove block from mold.

55
Sectioning

1. With a sharp razor blade, block the molded paraffin on all four edges,
thus paring down volume of block. Make sure the edges are straight and
that the ends of the rectangle are perpendicular to the sides.

2. Place block in rotary microtome chuck; fasten securely.

3. Orient blade flush with block surface.

4. Section at 15um until tissue is reached. When tissue is reached,
section a 6pm.

5. Lay ribbons of sections on brown paper in a tray.

6. Mount sections by floating them in pairs in water (52-56OC) and
lifting them off with gelatinized slides.

7. Allow sections to dry vertically.

8. Deparaffinize in 4 changes of xylene, 20 minutes each.

8. Hydrate according to following schedule.

100% alcohol 10 minutes
95% alcohol 10 minutes
70% alcohol 20 minutes
50% alcohol 20 minutes
Distilled water 20 minutes
10. Stain.
Staining

1. Place slides in 1% acqueous thionine (cold) for 10 minutes.
Thionine
200 m1 of 1M NaOH
764 m1 of distilled water
10 g of thionine
Must be mixed and heated for one hour at 900 C
2. Rinse briefly in distilled water.
3. Place sections in 70% alcohol for 5 minutes.
4. Differentiate in 95% alcohol. (This happens very fast.)
5. Place in 100% for 5 minutes.
6. Clear in Xylene for 20 minutes.

7. Cover slip.

APPENDIX C:

Raw Data

Plasma Corticosterone Concentrations

56

(ug/100ml plasma)

 

 

 

 

Group Animal if Value E
0600 D1 12.5

D2 20.0

D3 16.5 17.6

D4 21.5
1200 l p;m. 1 14.5

1 p.m. 2 13.0

1 p.m. 3 30.0

1 p.m. 4 40.5 24.5
1800 N1 41.0

N2 81.0

N3 60.0

N;__ 29.5 52.9
2400 l a.m. 1 15.5

1 a.m. 2 12.5

1 a.m. 3 14.5

gka.m. 4 12.0 13.6

 

57

Multiple and marginated nucleoli
PVm

 

 

 

 

 

 

 

 

Gronp, Animal # % multiples % magggnation
plane singles multiples
0600 D1 (coronal) 9.9 25.7 86.7
D2 (horizontal) 16.2 24.0 96.0
D3 (coronal) 23.2 25.0 91.4
D4 (horizontal) 15.4 19.8 95.7
I 16.2 23.6 92.5
1200 l p.m. l (coronal) 23.6 22.1 97.1
1 p.m. 2 (horizontal) 20.9 15.7 93.8
1 p.m. 3 (coronal) 15.3 25.4 95.5
1 p.m. 4__(horizontal) 37.9 20.7 92.5
E 24.2 21.0 94.7
1800 N1 (coronal) 16.2 34.4 86.9
N2 (horizontal) 17.0 18.9 84.6
N3 (coronal) 10.5 32.6 68.8
N4 (horizontal) 29.1 26.2 88.6
E 18.2 28.0 82.2
2400 1 a.m. l (coronal) 18.8 16.0 75.9
1 a.m. 2 (coronal) 25.2 31.6 96.9
1 a.m. 3 (horizontal) 14.6 18.6 95.4
1 a.m. 4 (horizontal) 14.4 11.4 72.7
'i' 18.3 19.4 85.2

 

Multiple and Marginated Nucleoli

58

 

 

 

 

 

 

 

 

PVl
Group Animal # % miltiples % maggination
plane singles multiples
0600 D1 (coronal) 18.4 41.9 96.4
D2 (horizontal) 34.4 29.7 86.8
D3 (coronal) 41.4 36.0 88.9
D4 (horizontal) 33.1 31.7 90.0
3? 31.8 34.8 90.5
1200 l p.m. l (coronal) 24.0 42.1 91.7
1 p.m. 2 (horizontal) 40.8 35.6 84.5
1 p.m. 3 (coronal) 13.5 27.4 90.5
1 p.m. 4 (horizontal) 42.9 31.8 81.3
I 30.3 34.2 87.0
1800 N1 (coronal) , 32.0 36.3 93.8
N2 (horizontal) 27.3 25.0 81.0
N3 (coronal) 16.3 36.7 96.0
N4 (horizontal) 41.3 34.1 85.9
E 29.2 33.0 89.2
2400 l a.m. l (coronal) 26.6 28.3 70.7
1 a.m. 2 (coronal) 32.2 46.6 91.8
1 a.m. 3 (horizontal) 25.3 32.2 82.1
1 a.m. 4 (horizontal) 15.9 20.5 62.5
E 25.0 31.9 76.8

 

59

Multiple and Marginated Nucleoli
SOa

 

 

 

 

 

 

 

 

Group Animal # % multiples % mapg2nation
plane singles multiples
0600 D1 (coronal) 14.6 37.2 93.4
D2 (horizontal) 12.6 34.1 89.5
D3 (coronal) 17.9 37.9 100.0
D4 (horizontal) 9.7 29.3 90.9
E 13.7 34.6 93.4
1200 l p.m. 1 (coronal) 23.2 33.3 92.1
1 p.m. 2 (horizontal) 26.0 21.7 87.2
1 p.m. 3 (coronal) 14.6 32.9 100.0
1 p.m. 4 ,(horizontal)_ 38.2 33.0 91.4
3? 25.2 30.2 92.7
1800 N1 (coronal) _ 25.5 36.8 94.9
N2 (horizontal) 37.6 27.6 91.5
N3 (coronal) 29.6 29.0 86.7
N4 (horizontal) 32.6 20.8 83.7
2400 l a.m. 1 (coronal) 20.1 34.9 80.6
1 a.m. 2 (coronal) 19.2 30.3 96.6
1 a.m. 3 (horizontal) 19.6 27.6 76.7
_2Ep,m. 4 (horizontal) 6.0 11.3 66.7
E’ 16.2 26.0 80.2

 

60
Multiple and Marginated Nucleoli

 

 

 

 

 

 

 

 

 

 

SCdm
2!ppp_ Animal # % multiples % maggination
singles ppgggpgfg;
plane invg.* ninvg,** inyg, ninvg. invg. ninvg.
0600 D1 (coronal) 25.4 46 85.1 78.1 100 93.1
. D2 (horizontal) 51.6 42.9 86.7 58.3 100 100
D3 (coronal) 37.5 38.5 88.9 85 97 100
D4 (horizontal) 31.7 17.5 86.6 63.8 100 100
E 36.6 36.2 86.8 71.4 99.3 98.3
1200 1 p.m. l (coronal) 28.5 28.2 86.4 71.4 94.4 100
l p.m. 2 (horizontal)33.6 40.4 86.3 60.7 100 89.5
1 p.m. 3 (coronal) 31.8 31.8 84.4 81 100 92.6
1 p.mm 4_(horizontal)37.4 40.7 74.4 60 100 100
SE 32.8 35.3 82.9 68.3 98.6 95.5
1800 N1 (coronal) 42.7 30.9 85.5 78.9 95 82.3
N2 (horizontal) 30.4 27.9 68.3 46.3 96.6 73.1
N3 (coronal) 49.5 42.9 93.6 75 100 100
N4 (horizontal) 44.1 32.1 73.7 83.3 97.8 82.4
E 41.7 33.5 Joe 70.9 97.4 84.5
2400 1 a.m.l(corona1) 28.1 23.2 78.2 60.5 100 100
1 a.m. 2 (coronal) 43.7 44 95 83.3 100 97.4
1 a.m. 3 (horizortal) 36.1 39.5 71.7 59.2 100 78.1
a.m..4 (horizontal) 47.4 38.3 72 37.8 97.8 91.3
2 'E _ x 38.8 , 36.3 79.2 60.2 99.5 91.7

 

* invg. - cells with an invaginated nucleus
** ninvg.- cells without an invaginated nucleus

61

Multiple and Marginated Nucleoli
SCvl

 

 

 

 

 

 

 

 

 

 

saggy; .Animal # % multiples % mapglnation
singles multiples
plane inyg,_ nipyg, inyg, ninyg, invg, ninvg,
0600' D1 (coronal) 22.7 27.6 86.2 74.5 100 95.2
D2 (horizontal)37.1 31.3 83.6 50 100 90
D3 (coronal) 41 48.1 86.4 67.9 100 100
D4 (horizontal)33 23.5 78.8 47.3 100 75
I 33.5 32.6 83.8 59.9 100 90.1
1200 l p.m. 1 (coronal) 29.1 26.9 86.9 47.4 92 100
l p.m. 2 (horizontal)27.3 25.7 77.5 53.8 96.7 100
1 p.m. 3 (coronal) 26.6 24.1 85.8 70.7 100 92.3
1 p,m. 4 (horizontal)29 39.6 85.7 86.2 100 94.7
3? 28 29.1 84 64.5 97.2 96.8
1800 N1 (coronal) 29.3 28 90 77.8 100 92.9
N2 (horizontal)l7.2 32.3 69.3 33.3 90.5 50
N3 boronal) 49.1 59.5 90.2 64.7 100 96
N4 (horizontal) 56.8 16.7 79.2 56 90.5 80
3? 38.1 34.1 82.2 58 95.3 79.7
2400 l a.m. l (coronal) 31.4 22.1 75 59 100 100
l a.m. 2 (coronal) 54.3 50 92.9 66.7 100 90
1 arm. 3 (horizontal)29.8 26.3 88.2 42.9 100 90
1 a.m. 4 (horizontal)32.5 39.5 71.2 43 97.5 86.7
'3? 37 34.5 81.8 52.9 99.4 91.7

 

62

 

 

 

 

 

 

 

Cell Areas (umz)
2.312 ”Animals 22.23121125224232le2
plane

0600 D1 (coronal) 154.0 161.2 138.9 32.8 40.1 222.7
D2 (horizontal) 195.5 217.0 170.9 70.2 74.6 214.7

03 (coronal) 126.9 130.9 135.7 35.9 43.2 243.9

D4 (horizontal) 179.4 206.6 156.9 52.3 64.3 192.1

'§' 164.0 179.0 150.6 47.8 55.5 218.4

1200 1 p.m. l (coronal) 185.1 146.3 119.1 30.4 36.3 194.3
1 p.m. 2 (horizontal) 159.2 186.7 138.5 50.1 55.9 193.2

1 p.m. 3 (coronal) 165.7 194.5 132.3 32.3 40.4 232.1

1 ppm. 4 (horizontal) 168.0 169.8 128.5 48.0 55.7 171.2

'E’ 169.5 174.4 129.6 40.2 47.1 197.7

1800 N1 (coronal) 225.6. 251.8 173.4 43.6 50.9 304.9
N2 (horizontal) 213.5 207.8 202.6 79.6 84.1 239.9

N3 (coronal) 246.4 292.9 167.9 41.5 45.7 292.4

N4 (horizontal) 185.3 327.7 192.4 66.6 71.7 229.6

'i' 217.7 270.0 184.1 57.8 63.1 266.7

2400 l a.m. l (coronal) 197.3 212.6 226.6 37.9 51.2 240.9
1 a.m. 2 (coronal) 175.7 260.8 160.6 33.1 43.2 245.6

1 a.m. 3 (horizontal) 212.4 217.4 227.3 71.2 84.2 184.7

1 a.m. 4 (horizontal) 238.4 244.5 232.5 62.1 77.3 215.2

'§' 205.9 233.8 211.8 51.1 64.0 221.6

 

MICHIGAN STATE UNIVERSITY LIBRARIES

0

H ll‘ll Jammy!

3 1293