—"'w'

 

\ u;,v‘

Ilm M1 H mmmmjmmmmml 1.

31293 00 50 0161

LIBRARY
Michigan State
University

 

This is to certify that the
dissertation entitled
HYPOTHALAMIC NEURAL PATHWAYS THAT MEDIATE

PHOTOPERIODIC CONTROL OF REPRODUCTION IN THE SYRIAN
HAMSTER (MESOCRICETUS AURATUS)

presented by

Michael Howard Brown

has been accepted towards fulfillment
of the requirements for

Doctor of Philosophy degreeinNeuroscience/Psychology

Cur—71;;

Major profess?» /

Date October 28, 1986

"(Iii-.- ‘m——— ’ ‘

0-12771

 

 

MSU

 

 

 

RETURNING MATERIALS:
Place in book drop to

 

 

LlfiRARJES remove this checkout from
w your record. w§ will
be charged if book is
returned after the date
stamped below.
~ My

' '0I994

 

 

 

HYPOTHALAMIC NEURAL PATHWAYS THAT MEDIATE PHOTOPERIODIC
CONTROL OF REPRODUCTION IN THE SYRIAN HAMSTER W

W

By

Michael Howard Brown

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology and Neuroscience Program

1986

ABSTRACT

HYPOTHALAMIC NEURAL PATHWAYS THAT MEDIATE PHOTOPERIODIC
CONTROL OF REPRODUCTION IN THE SYRIAN HAMSTER W

AllRAIllS.)
By
Michael Howard Brown

Exposure of male golden hamsters to photoperiods providing less than 12.5 hours of
light per day results in testicular regression within 8-12 weeks. This response is thought to
be mediated by a multisynaptic neural pathway by which photic information is relayed
from the retina through the suprachiasmatic hypothalamic nucleus (SCN), paraventricular
hypothalamic nucleus (PVN), preganglionic sympathetic neurons in the spinal cord, and
postganglionic sympathetic neurons of the superior cervical ganglion (SGC) to the pineal
gland Anatomical evidence for these pathways has been provided and ablation of the
SCN, the PVN, the SCG, or the pineal prevents testicular regression in hamsters exposed
to short days. The present investigation was designed to further explore the hypothalamic
circuits involved in this phenomenon utilizing hypothalamic damage less severe than that
employed by previous researchers.

Bilateral knife cuts placed in a horizontal plane ventral to, through, or dorsal to the
PVN all prevented testicular regression in hamsters exposed to short days but affected
neither the circadian pattern of wheel-running activity nor the phase relationship between
this behavior and the light-dark cycle. Cuts that either produced unilateral damage or were
anterior to the PVN had no effect on the testicular response to short days.

Other investigators have provided anatomical evidence that some paraventriculo—spinal
fibers take a lateral course upon leaving the PVN while others exit the nucleus via a dorsal
route. In the present investigation, injections of horseradish peroxidase (I-IRP) into the
spinal cord of hamsters that received cuts dorsal to the PVN, that either were bilateral and

prevented gonadal regression or were intentionally made on only one side, resulted in

labelled neurons in the PVN. Thus, while cuts dorsal to the PVN probably sever some
paraventriculo-spinal axons and this may be the reason for their efficacy in abolishing
photoperiodic responses, it appears that destruction of only a portion of these connections
is sufficient to produce this effect.

Either knife cuts (present investigation) or electrolytic lesions (previous reports) in the
area of the PVN damage, in addition to neurons of the PVN and their axons, fibers of
possible relevance to photoperiodism that pass through the area. Therefore,
intrahypothalamic injections of N-methyl-D,L-aspartate, a toxin that destroys some
neurons but is thought not to damage fibers passing through the injection site, were used
to test the possibility that the effect of the knife cuts and electrolytic lesions were due to
damage to these fibers of passage. Unilateral injections of NMA near the PVN destroyed
most of the neurons of the posterior portions of the PVN, and produced a statistically
significant reduction in the number of PVN neurons labelled after injections of I—IRP into
the spinal cord. Bilateral injections of NMA that were centered in or very near the PVN
also produced substantial damage to the posterior part of that nucleus, and furthermore,
prevented the testicular response to short days. Five of these animals received tests of
sexual behavior and 4 of them were able to inseminate hormonally-primed, sexually
receptive females. Injections of NMA that were centered more distant from the PVN
produced either no, or minor damage to that nucleus and either no, or partial effects on the
testicular response to short days, respectively. That injections ineffective in abolishing
testicular regression produced little or no damage to that portion of the hypothalamo-
hypophyseal axis responsible for the secretion of luteinizing hormone was suggested by
the fact that the 10 animals that were exposed to short days for 28 weeks all exhibited
complete spontaneous testicular recrudescence.

These results support the hypothesis that neural projections from the SCN to the PVN,
and projections from the PVN to the spinal cord mediate gonadal regression in hamsters

exposed to short days. Furthermore, disruption of only a portion of the paraventn‘culo-

spinal projections is sufficient to prevent photoperiod-dependent gonadal regression.

ACKNOWLEDGEMENTS

For all of her help with Experiment 3 and two "follow-up" experiments. Lori Badura
has practically earned a coauthorship. I'm sorry I couldn't give it to you, Lori. The
graduate office won't accept coauthored dissertations, but thanks for all of your help.

Beth Dettmer-Radtke did a great job of keeping me supplied with chemical solutions
and other necessities. Mark Griswald spent many long hours helping me mount tissue.
The two of them did a great deal of work on two of my "side projects" while I was intently
collecting data for the dissertation.

Ken Smithson, Kevin Grant, and Tim Youngstrom answered a multitude of my naive
questions about photography and photomicrography.

The members of my dissertation and guideance committee, Drs. Glenn Hatton, John
Johnson, and Cheryl Sisk, have provided many valuable comments and criticisms
regarding the manuscript, the experiments, and neuroscience in general. Cheryl was lcind
enough to take the time to read, and comment, on a portion of the text while she was
frantically preparing her grant proposal. I especially want to thank Glenn, and the
Neuroscince Program for the use of much of the equipment that made this work possible.

Tony Nunez chaired my MA and PhD committies, Gave me free-run of the lab and
virtually unlimited use of his grant funding, let me test my own ideas-even some of the
wildest of them, and literally got me started in "this business" (as he himself would say).

Thanks to all of you.

This research was funded by NIH grant MH 37877 and A.U.R.I.G. funds to Tony

Nunez.

ii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. v
LIST OF FIGURES .............................................................................. vi
INTRODUCTION ................................................................................. 1
EXPERIMENT 1 .................................................................................. 5
Methods ..................................................................................... 6
Results ...................................................................................... 9
Discussion ................................................................................. 13
EXPERIMENT 2 ................................................................................. 17

Part A: Do Knife Cuts Dorsal to the PVN Interrupt Paraventriculo—Spinal
Projections? ............................................................................... 19

Part B: Paraventriculo—Spinal Projections and the Testicular Response to

Short Days ................................................................................ 31
Methods ................................................................................... 32
Results ..................................................................................... 33
Discussion ................................................................................. 39
EXPERIMENT 3 ................................................................................. 47

Part A: Are Neurons of the Paraventricular Nucleus Sensitive to the Toxic

Effects of N-Methylaspartate? .......................................................... 48
Methods ................................................................................... 49
Results ..................................................................................... 50
Discussion ................................................................................. 56

iii

Part B: Axon-Sparing Lesions of the Paraventricular NucleuszConfirmation

of the Role of the PVN in PhotOperiod-Dependent Seasonal

Reproductive Cycles ..................................................................... 57
Methods ................................................................................... 57
Results ..................................................................................... 63
Discussion ................................................................................. 82
GENERAL DISCUSSION ...................................................................... 91
LIST OF REFERENCES ........................................................................ 95

iv

LIST OF TABLES

Table 1: Number of hamsters sacrificed at each of three times and number of hamsters

that were used in the sexual behavior tests ....................................................... 61

Table 2: Testicular and seminal vesicle weights ................................................. 7 O

LIST OF FIGURES

Figure 1: Schematic representation of knife cuts that produced bilateral damage to
the hypothalamus. The cuts (dashed lines) are shown at the point of greatest bilateral
damage for each case using drawings (A and B) modified from those

published by Lehman et a1. (1984). The individual cases are identified by numbers.
Abbreviations are: 0C, optic chiasm; OT, optic tract; PVN, paraventricular

nucleus; SCN, suprachiasmatic nucleus; ZI, zona incerta ..................................... 10

Figure 2: Photornicrograph of a coronal section (50 um thick, pyronin Y stain)
through the hypothalamus of a hamster (No. 1) in which the knife cut (indicated by
arrows) produced bilateral damage just dorsal to the PVN. For abbreviations, see

Figure 1 .............................................................................................. 11

Figure 3: Mean (iSEM) testicular weights of the short-day housed hamsters with
bilateral knife cuts (n = 7) and sham-operated hamsters kept in short days (shaded
bar; n = 9) or long days (open bar; n = 6) ....................................................... 12

Figure 4: Activity record of a hamster (No. 2) that received a bilateral knife cut

ventral to the PVN and did not show testicular regression after 13 weeks of

exposure to short days. The black bar at the top represents the dark portion of the
light-dark cycle ...................................................................................... 14

Figure 5: Location of the area (indicated by batching) of the lateral hypothalamus in
which counts of HRP labelled cells were made. The drawings (A-C, rostral to

caudal) of the hamster hypothalamus are a modification of those published by

vi

Malsbury (1977). Abbreviations are: ARC, arcuate nucleus; F, fomix; OT, optic

tract; VMN, ventromedial nucleus ................................................................ 20

Figure 6: Schematic representation of the distribution of HRP-labelled neurons in
three sections (at 240 um intervals, A-C are rostral to caudal) through the
paraventricular nucleus of a hamster (No. 122) that received a unilateral knife cut
(indicated by arrows) before bilateral injections of HRP into the spinal cord.
Note the reduction in the number of HRP-labelled cells ipsilateral to the cut
Abbreviations are: 0C, optic chiasm; OT, optic tract; PVN, paraventricular
nucleus; SCN suprachiasmatic nucleus; ZI, zona incerta; 3V, third ventricle.

Each dot represents one labelled cell ............................................................. 22

Figure 7: Schematic representation of the distribution of HRP-labelled neurons in

three sections (at 240 um intervals, A-C are rostral to caudal) through the

paraventricular nucleus (PVN) of a hamster (N o. 128) that received a unilateral

knife cut (indicated by arrows) before bilateral injections of HRP into the spinal

cord. Note that the cut was more dorsal and centered more anterior than that in

hamster No. 122 (Figure 5). Note also that the reduction in the number of HRP-

labelled cells ipsilateral to the cut was less substantial than that in hamster No. 122 ....... 25

Figure 8: Schematic representation of the distribution of HRP-labelled neurons in

three sections (at 240 um intervals; A-C, rostral to caudal) through the

paraventricular nucleus (PVN) of a hamster (N o. 123) that received a unilateral

knife cut (indicated by arrows) before bilateral injections of HRP into the spinal

cord. Note that the cut was dorsal to the PVN and failed to produce a reduction in

the number of HRP-labelled cells in the PVN ipsilateral to the cut. .......................... 27

Figure 9: Testicular widths of individual animals in which the knife cuts failed to
produce the intended bilateral damage dorsal to or through the PVN (solid lines) and

vii

mean testicular width of control hamsters (n = 5) exposed to short days (broken line)... .35

Figure 10: Schematic representation of the location of knife cuts that produced

bilateral damage to the hypothalamus in Experiment 2. The cuts are drawn to show

their location and extent in sections (A or B) through the PVN although in most

cases the cuts were centered anterior to the nucleus. In one hamster (No. 103, not

shown, the cuts produced a bilateral mechanical lesion of the PVN). Abbreviations

are as in Figure 1 ................................................................................... 36

Figure 11: Mean testicular width for the 7 hamsters with bilateral knife cuts (KC)
through or dorsal to the PVN, for 5 sham-operated control hamsters that were
housed in short-day photoperiod, and 5 sham-operated hamsters that were

housed in long-day photoperiod. Data for one hamster (No. 103) in which the
knife cut produced a mechanical lesion of the PVN are included with data for the
other hamsters that received knife cuts. Two of the hamsters that received knife cuts
were sacrificed early, hence, the last two data points (weeks 14 and 16) do not

include data for them ............................................................................... 37

Figure 12: Schematic representation of the distribution of HRP-labelled neurons in

three sections (at 240 um intervals; A-C, rostral to caudal) through the

paraventricular nucleus (PVN) of a hamster (No. 115) that received a bilateral knife

cut (indicated by arrows) that prevented testicular regression. This animal received
bilateral injections of HRP into the spinal cord and was sacrificed 48 hr later. Note

that the cut was dorsal to the PVN and failed to prevent labelling of cells in the PVN.. . . .40

Figure 13: Schematic representation of the distribution of HRP-labelled neurons in
three sections (at 240 um intervals; A-C, rostral to caudal) through the
paraventricular nucleus (PVN) of a hamster (No. 118) that had received a bilateral

knife cut (indicated by arrows) that prevented testicular regression before receiving

viii

bilateral injections of HRP into the spinal cord. Note that the cut was dorsal to the
PVN and failed to prevent labelling of cells in the PVN ....................................... 42

Figure 14: Darkfield photomicrograph of a section (30 pm thick, counterstained

with pyronin Y) through the hypothalamus of a hamster (No. 118) in which a

bilateral knife cut (indicated by large arrows) just dorsal to the PVN prevented

testicular regression but did not prevent labelling of PVN neurons (indicated by

small arrows) after injections of HRP into the spinal cord ..................................... 44

Figure 15: Schematic representation of the locations of the centers of the unilateral
injections of NMA in 4 hamsters that received bilateral injections of HRP into the
spinal cord after the NMA injections. The drawings of the hamster hypothalamus
(A-C are rostral to caudal) are modifications of those published by Malsbury
(1977). Abbreviations are: ARC, arcuate nucleus; F, fomix; OT, optic tract; PV,

paraventricular nucleus; SC, suprachiasmatic nucleus; SO, supraoptic nucleus ............ 51

Figure 16: Photomicrographs of a series of sections (30 um thick, counterstained

with pyronin Y) at 180 um intervals (A-D, rostral to caudal) through the PVN of

a hamster (No. 380) that received a unilateral injection of NMA just dorsolateral

to the PVN prior to bilateral injections of HRP into the spinal cord. Note that in

the most anterior section (A) there does not appear to be any loss of unlabelled

cells on the injected (left) side; however, the number of both HRP-labelled, and
unlabelled cells on the injected side is substantially reduced in the more posterior

sections (B-D) ...................................................................................... .53

Figure 17: Mean (iSEM) number of HRP-labelled neurons in the PVN ipsilateral
and contralateral to the site of injection of NMA. The centers of the sites of the

NMA injections in the four hamsters were represented in Figure 15 ......................... 55

Figure 18: Mean testicular width of 13 control hamsters that were housed in short
ix

days after receiving injections of ACSF, 5 unoperated control hamsters that were
housed in long days, and 5 hamsters that were housed in short days after receiving

injections of NMA that caused "complete" damage to the PVN ............................... 64

Figure 19: Schematic representation of the locations of the centers of the injections
of NMA that produced "complete" bilateral destruction of the PVN. The pairs of
circled numbers represent the bilateral injection sites in individual hamsters. The
drawings of the hamster hypothalamus (A-B, rostral to caudal) are modifications
of those published by Malsbury (1977). Abbreviations are: ARC, arcuate nucleus;
F, fomix; IC, internal capsule; OT, optic tract; MAH, medial anterior
hypothalamus; MFB, medial forebrain bundle; MT, mammillo—thalamic tract; PV,

paraventricular nucleus; RT, reticular thalamic nucleus; V, ventral thalamic nucleus ....... 6 5

Figure 20: Photomicrographs of a series of 40 um thick sections (160 um intervals,
cresyl violet stain) through the hypothalamus of a hamster (No. 3) representative of
those cases in which the loss of neurons of the PVN, following bilateral injections
of NMA, was judged to be "complete" damage. Notice that while PVN neurons are
visible in the more anterior sections (A and B), there was a substantial loss of
neurons from the posterior PVN (C and D). The arrows in D indicate the gliosis

surrounding the needle tracks ..................................................................... 67

Figure 21: Locations of the centers of the injections of NMA that produced partial
damage to the PVN. The drawings of the hamster hypothalamus (A-D, rostral to
caudal) are modifications of those published by Malsbury (1977). Abbreviations
are: P, fomix; IC, internal capsule; OT, optic tract; MAH, medial anterior
hypothalamus; MFB, medial forebrain bundle; PV, paraventricular nucleus; RT,
reticular thalamic nucleus; SC, suprachiasmatic nucleus; SO, supraoptic nucleus;

V, ventral thalamic nucleus ........................................................................ 71

Figure 22: Testicular widths of individual hamsters in which the injections of NMA
(presented in Figure 21) produced partial damage to the PVN ................................ 73

Figure 23: Locations of the centers of the injections of NMA that did not damage
the PVN or prevent testicular regression in hamsters that were exposed to short
days for 14 weeks. The drawings of the hamster hypothalamus (A-E, rostral to
caudal) are modifications of those published by Malsbury (1977). Abbreviations
are: AC, anterior commissure; F, fomix; IC, internal capsule; OC, optic chiasm;
OT, optic tract; MAH, medial anterior hypothalamus; MFB, medial forebrain
bundle; MPO, medial preoptic area; PV, paraventricular nucleus; RT, reticular
thalamic nucleus; SC, suprachiasmatic nucleus; SM, stria medullaris; SO,

supraoptic nucleus; V, ventral thalamic nucleus ................................................. 7 5

Figure 24: Locations of the centers of the injections of NMA that did not damage
the PVN of hamsters that were exposed to short days for 25-28 weeks. The
drawings of the hamster hypothalamus (A-E, rostral to caudal) are modifications of
those published by Malsbury (1977). Abbreviations are: AC, anterior commissure;
F, fomix; IC, internal capsule; OC, optic chiasm; OT, optic tract; MAH, medial
anterior hypothalamus; MFB, medial forebrain bundle; MPO, medial preoptic area;
PV, paraventricular nucleus; RT, reticular thalamic nucleus; SC, suprachiasmatic

nucleus; SM, stria medullaris; SO, supraoptic nucleus; V, ventral thalamic nucleus ....... 77

Figure 25: Photomicrograph of a section (40 um thick, cresyl violet stain) through
the brain of a hamster in which the NMA injections were centered in the thalamus
approximately 750 um dorsal to the PVN. The gliosis and loss of neurons is
typical of cases in which the NMA injections were centered in the thalamus. Note
that the damage did not extend to the PVN. The open arrow indicates the zone of

dense gliosis surrounding the center of the injection site ....................................... 80

xi

Figure 26: Testicular widths of the 9 hamsters for which the injection sites were
presented in Figure 24 and for 4 control hamsters that were exposed to short days

for 25- 28 weeks .................................................................................... 81

Figure 27: Mean number of mounts (without intromissions) and latency to mount
exhibited by hamsters that received injections of NMA (n = 5) that caused either
complete (11 = 4) or partial (n =1) damage to the PVN, and unoperated control
hamsters that were housed in long days. While the number of mounts did not
differ statistically between groups, the difference in latency to mount was

statistically significant (p < 0.05) ................................................................. 83

Figure 28: Mean number of intromissions and latency to intromit exhibited by
hamsters that received injections of NMA (n = 5) that caused either complete
(n = 4) or partial (n =1) damage to the PVN, and unoperated control hamsters
that were housed in long days. Statistically significant differences between
groups were obtained both for number of intromissions (p < 0.01) and for

latency to intromit (p < 0.05) ...................................................................... 84

xii

INTRODUCTION

Many mammalian species are photoperiodic. These animals show seasonal
reproductive cycles in which reproductive competence is dependent on the length of the
light portion of the li ght-dark cycle. In a long-day breeder such as the golden hamster
W5), adult gonadal development is maintained while the animal is
exposed to a stimulatory, long-day photoperiod (i.e. >12.5 hr of light per 24 hr in the case
of the golden hamster). When golden hamsters are exposed to nonstimulatory (short-day)
photoperiods, the hypothalamo-hypophysial axis, through a mechanism involving pineal
melatonin secretion, becomes hypersensitive to negative feedback from gonadal steroids
(T amarkin, et al., 1976). This leads to a reduction in the release of gonadotmpins and
regression of the gonads to a prepubertal state (i.e. gametogenesis ceases, steroid release
decreases, and, in males, the testes show reduced size while, in females, ovaries exhibit
proliferation of interstitial tissue and lack of corpora lutea (for review-see Reiter, et al.,
1983). In male golden hamsters, 8-12 weeks of exposure to nonstimulatory photoperiod is
generally considered to be sufficient for complete gonadal regression. Prolongation of
exposure to nonstimulatory photoperiod after the occurrence of gonadal regression results
in photorefractoriness followed by spontaneous reproductive recrudescence. Testicular
recrudescence is preceded by increases in circulating titers of gonadotropins and prolactin
and a reduction in sensitivity of the hypothalamo-hypophysial axis to negative feedback
from gonadal steroids (Ellis and Turek, 1979; Matt and Stetson, 1979; Turek and Ellis,
1981) leading to spontaneous gonadal recrudescence which begins after 15-20 weeks of
exposure to nonstimulatory photoperiod (Steger, et al., 1985). Hamsters exhibiting

photorefractoriness fail to show gonadal regression upon further exposure to
1

2

nonstimulatory photoperiod but will regain sensitivity to photoperiod if exposed to a
stimulatory photoperiod for 7-11 weeks (Stetson, et al., 1977).

The neural pathways that mediate this response have long been debated, the proposed
circuit having been modified several times. The current model involves a long circuitous
pathway begining at the retina. Some retinal ganglion cells send axons into the
hypothalamus which terminate in the suprachiasmatic nuclei (SCN; Moore and Lenn,
1972; Pickard and Silverman, 1981), forming the retinohypothalamic tract. The SCN
contain a circadian oscillator (for review, see Rusak and Zucker, 1979) and appear to be
necessary for seasonal reproductive cycles since SCN lesions prevent gonadal regression
in hamsters housed in short-day photoperiod (Rusak and Morin, 1976). The SCN in turn
send axonal projections to the paraventricular nucleus (PVN) of the hypothalamus (Berk
and Finkelstein, 1981; Silverman, et al., 1981; Stephan, et al., 1981). Lesions of the PVN
prevent gonadal regression in hamsters exposed to nonstimulatory photoperiods
(Bartness, et al., 1985; Lehman, et al., 1984; Pickard and Turek, 1984) as well as the
nighttime rise in pineal melatonin production both in rats (Klein, et al., 1983) and
hamsters (Lehman, et al., 1984). Some of the cells of the PVN send long descending
fibers to the interrnediolateral cell column of the spinal cord (Armstrong, et al., 1980; Don
Carlos and Finkelstein, 1982; Swanson and Kuypers, 1980b). Among the areas
innervated by these axons are the upper thoracic segments (Tl-T3) which contain
preganglionic sympathetic cell bodies that innervate the superior cervical ganglion (Rando,
et al., 1981). Efferent fibers of the superior cervical ganglion (SCG) form the nervi conarii
which provide sympathetic input to the pineal gland (Ariens Kappers, 1960). Interruption
of the pathway by superior cervical ganglionectomy, central deafferentation of the SCG,
or pinealectomy (Reiter and Hester, 1966) results in insensitivity to photoperiod as
evidenced by lack of gonadal regression in hamsters exposed to short-day photoperiods.

Evidence implicating the PVN in relaying photoperiodic information to the spinal cord
should not be regarded as conclusive yet. Among the efferent projections of the SCN are

3

fibers that course dorsally through and by the PVN (Stephan, et al., 1981). Some of these
fibers continue coursing dorsally through the paraventricular thalamic nucleus to the
habenula. A subset of these axons continues caudally through the mesencephalic
periaqueductal grey, travelling at least as far caudally as the dorsal raphe nucleus. The
habenula in turn, projects to the pineal (Dafny, 1983). The PVN also sends axons to the
habenula (Conrad and Pfaff, 1976) as well as directly to the pineal (Buijs and Pevet, 1980;
Guerillot, et al., 1982; Korf and Wagner, 1980; Moller and Korf, 1983). All of these
pathways are disrupted by the electrolytic lesions of the PVN that have been used to
abolish circadian rhythms of pineal serotonin N-acetyltransferase (NAT) activity and
melatonin content in rats (Klein, et al., 1983), and gonadal regression in hamsters exposed
to nonstimulatory photoperiod (Bartness, et al., 1985; Lehman, et al., 1984; Pickard and
Turek, 1983). These direct projections from the brain to the pineal gland are not sufficient
to maintain photoperiodically mediated seasonal reproductive cycles since superior cervical
ganglionectomy abolishes the cycles (Reiter and Hester, 1966). Nevertheless, these
alternative pathways may be necessary for normal circadian rhythms in the pineal and for
gonadal regression in hamsters housed in short-day photoperiod.

Additionally, other hypothalamic areas send descending projections to autonomic
regions of the brainstem and spinal cord (Don Carlos and Finkelstein, 1982; Swanson and
Kuypers, 1980a). These other areas may also be involved in neural regulation of pineal
function and photoperiodic reproductive cycles. A possibility that has not yet been ruled
out is that the efficacy of PVN lesions may be due to interruption of fibers (from the SCN)
passing through the lesion site to other areas.

To further explore the hypothalamic neural pathways involved in photoperiodic
control of reproduction, the following experiments were performed. The specific goal of
the research was to confirm the role of the PVN in photoperiodic regulation of
reproductive cycles through the use of hypothalamic knife cuts, intrahypothalamic

injections of neurotoxin, and neural tract tracing by the use of horseradish peroxidase

4

(HRP). Experiment 1 was designed to examine the role of fibers that project from the
SCN to the PVN in testicular regression via placement of knife cuts between these two
nuclei. In Experiment 2, knife cuts placed in the path of paraventriculo-spinal axons and
injections of HRP into the spinal cord were used to investigate the role of projections from
the PVN to the spinal cord in testicular regression. Intrahypothalamic injections of N-
methyl aspartate, a neurotoxin thought to destroy neuronal cell bodies without affecting
axons that pass through the injection site, were employed in Experiment 3 to test whether
fibers that project dorsally from the SCN through the PVN are involved in testicular

regression.

EXPERIMENT 1

As previously mentioned, the SCN contain a circadian oscillator (Rusak and Zucker,
1979) and send efferent projections to the PVN (Berk and Finkelstein, 1981; Silverman, et
al., 1981; Stephan, et al., 1981). Bilateral electrolytic lesions of either nucleus abolish
circadian rhythms of pineal NAT activity and melatonin content (Klein, et al., 1983;
Lehman, et al., 1984; Moore and Klein, 1974) and prevent testicular regression in
hamsters housed in short-day photoperiod (Bartness, et al., 1985; Lehman, et al., 1984;
Pickard and Turek, 1983; Rusak and Morin, 1976). These results suggest that neural input
to the PVN from the SCN is necessary for photoperiod-dependent reproductive cycles.

In the present experiment, knife cuts were placed in a horizontal plane aimed ventral
to the PVN before the hamsters were placed in a short-day photoperiod. Knife cuts placed
in the intended location should have interrupted projections from the SCN to the PVN
without damaging either nucleus. Thus, the experiment was carried out to test whether
fibers that project dorsally out of the SCN are necessary for testicular regression in
hamsters housed in short days. To verify that the retinohypothalamic tract was not
damaged by the knife cuts, some of the hamsters were given intraocular injections of
HRP. Additionally, circadian locomotor activity patterns of some of the animals were
monitored at the end of the experiment to test the ability of the animals to entrain to the
light-dark cycle.

Methods

Subjects and Housing

Adult male golden hamsters (LVG/LAK; Charles River Breeding Laboratory,
Newfield, NJ) were individually housed and given ad 1112 access to water and food. At the
begining of the experiment, animals were housed in a long-day photoperiod (16 hrs light:8
hrs dark). Light intensity was approximately 15 foot candles.
51112211

Twenty-five animals were assigned to one of two surgical conditions. Surgery was
performed under Equithesin (4.5 ml/kg body weight) anesthesia. For placement of knife
cuts, 10 hamsters were placed in a stereotaxic apparatus (Kopf Instruments) and an
incision was made through the skin on the dorsal surface of the skull. A piece of skull
(approximately 3 mm square) was removed, the superior sagittal sinus retracted, and a
Scouten (Scouten, et al., 1981) retractable wire knife (Kopf Instruments) was lowered
into the brain at the midline of the animal. Stereotaxic coordinates were 0.3 mm posterior
to the bregma, 5.8 mm ventral to the sinus with the top of the incisor bar 2.0 mm below
ear bar zero. With the cannula in position, the blade of the knife was extended 2 mm in a
direction perpendicular to the midline of the animal in order to puncture the ependymal
lining of the third ventricle. The knife was then rotated 90° counterclockwise (ie. until the
blade was parallel to the midline) and then 180° clockwise then returned to the original
position. With the outer guide cannula left in place, the blade was then retracted and the
procedure repeated on the opposite side of the animal. Finally, the knife was rotated a full
360° degrees in each direction before the blade was retracted and the knife removed from
the animal. Fifteen hamsters served as sham-operated control animals. These hamsters
received the incision to expose the skull but no neural damage.
Emcdm

Hamsters were returned to the long-day photoperiod and allowed to recover from

surgery for one week. The 10 hamsters with knife cuts and nine of the sham-operated

6

7

control hamsters were then moved to another colony room and housed in a short-day
photoperiod of 6 hrs light:18 hrs dark. The remaining six hamsters were kept in the
original long-day photoperiod. After the short-day housed hamsters had been in that
photoperiod for 13 weeks, the hamsters were again anesthetized with Equithesin and
castrated. Paired testicular weights for each animal were recorded to the nearest 0.1 mg.
Lon g-day housed control hamsters were sacrificed at the time of castration. Short-day
housed control hamsters and hamsters that received knife cuts were returned to the short-
day photoperiod and allowed to recover from surgery for two weeks. After that time, 3
control hamsters and 6 hamsters with knife cuts were placed in metal cages (21.5 x 11 x
11 cm) attached to activity wheels (35.5 cm in diameter x 14 cm wide). Photoperiod was
maintained at 6 hrs light: 18 hrs dark. Locomotor activity was monitored via optoelectronic
activity counters. Each counter consisted of an infrared light emitting diode aimed at a
phototransistor. The transistor and diode were placed in a position such that a cam on the
axle of the wheel interrupted the light beam once during each wheel revolution. The
phototransistors were connected to input channels of optoisolated input/output boards
(Mullen Computer Products) inside a 64K Northstar Horizon computer which also
contained a real-time clock board (Mountain Hardware, Inc.). Data were collected as
number of wheel revolutions per 12-min interval (i.e. 120 intervals per 24-hr day) and
stored on 5.25 in. floppy disks (Verbatim Corp.) every 24 hr. For visual presentation, the
data were collapsed into 36-min intervals. For each 36-min interval in which the number
of wheel revolutions exceeded 14, a dark rectangle was printed. A blank space was printed
for each interval in which 14 or fewer revolutions occurred. Data were printed on a 24~hr
time scale horizontally, with data for each day printed below data for the previous day.
Hamsters were allowed to habituate to the recording conditions for 15 days, after which
activity was monitored for 11 days. Visual inspection of the records was used to estimate
phase angle (i.e. time between onset of activity and onset of darkness) of the activity

patterns to the nearest 18 min.

Histology

After collection of locomotor activity data, eight of the hamsters that had received
knife cuts were anesthetized and given bilateral intraocular injections of 6 ul of a 30%
solution of HRP (Sigma Type VI). After a 24-hr survival period these animals were
sacrificed by overdose of sodium thiamylal (Surital; Parke-Davis) and perfused
transcardially with 5% saline containing Heparin (1000 U/l) followed by 4%
glutaraldehyde in 0.1M phosphate buffer (pH 7.4). The superphysiological concentration
of saline was selected in an attempt to reduce enlargement of the third ventricle during
histological processing of the tissue. Brains were removed and stored, refrigerated, for
48 hrs in 0.1M phosphate buffered sucrose solutions of increasing concentrations (5%,
10%, 15%, 20%; 6-12 hrs in each solution). In some cases a small amount of 4%
glutaraldehyde was added to the first two sucrose solutions to improve fixation of the
tissue. Frozen 50 um thick sections were cut in an oblique coronal plane to include the
SCN and the anterior part of the PVN in the same sections. Free-floating sections were
reacted with tetramethylbenzidine (TMB) according to published procedures (Mesularn, et
al., 1980). Alternate series of sections were mounted on chrome-alum subbed slides and
stained with either cresylecht violet or pyronin Y. Two hamsters (Nos. 4 and 9) died
before I-IRP injections were scheduled. Their brains were removed and stored for two
weeks in 10% formalin, then sectioned and stained with cresylecht violet. Bright- and
dark-field illumination at magnifications ranging from 25-100 X were used to verify the
presence of TMB reaction product in the SCN. Selected sections were traced with the aid
of a projecting microscope. From these tracings, schematic diagrams were made showing
the location of the cuts at the point of greatest bilateral damage. The tracings were made
without knowledge of the gonadal status of the individual animals. Finally, selected
sections were photographed using Kodak Technical Pan Film 2145 and prints were made
on Kodak F5 paper. Since the variance of testicular weight was higher for hamsters that

received knife cuts than for each of the two groups of control hamsters, two-tailed Mann-

9

Whitney 11 tests were used for statistical comparisons of testicular weights between the
three groups. In order to correct for multiple comparisons, the probability of incorrectly
stating that differences exist, or, was divided equally among each of the 3 possible
comparisons (Kirk, 1968). Thus, the critical value was 0.05/3 = 0.0166. In addition,

when feasible (i.e., when U was an intergral number), exact probabilities associated with

the obtained 11 values were reported either from tabled values (when n1 and n2 both < 9)
or (when either ml or n2 2 9) as estimated from the formula given by Kirk (1968). A two-

tailed Mann-Whitney U test was also used for comparison of the phase angle of the
activity patterns.
Results

In two hamsters (Nos. 6 and 7), the knife cuts produced unilateral damage to the
hypothalamus dorsal to the PVN. In one case (No. 12) no evidence of gliosis or neural
damage was detected. Seven hamsters had bilateral knife cuts that were placed in the
intended anterior-posterior location. As illustrated in Figure 1, however, the dorso—ventral
positioning of the knife cuts was quite variable. Knife cuts fell ventral to (No. 2), through
(Nos. 9, 10, 11), or slightly dorsal to (Nos. 1, 4, 5) the PVN. A photomicrograph of a
coronal section through the hypothalamus of a hamster (No. 1) representative of cases in
which the cuts were dorsal to the PVN is presented in Figure 2.

As summarized in Figure 3, all seven of the hamsters with confirmed bilateral knife
cuts had large testes (3.65 :t 0.44 g; mean :t SEM). The six sham-operated control
hamsters that were housed in long-day photoperiod also had large testes (4.07 :1: 0.20 g)
and did not differ significantly from animals with knife cuts on this measure (11 = 19,

p = 0.836). On the other hand, the nine sham-operated control hamsters housed in short-
day photoperiod had fully regressed testes (0.56 :1: 0.06 g) which were statistically
significantly different both from those of control hamsters that were housed in long days
(11 = 0, p = 0.00148) and from those of hamsters with bilateral knife cuts (11 = 0,

p = 0.00086). The three hamsters (Nos. 6, 7, and 12) in which knife cuts were either

10

 

 

 

 

 

Figure 1: Schematic representation of knife cuts that produced bilateral damage to the
hypothalamus. The cuts (dashed lines) are shown at the point of greatest bilateral damage
for each case using drawings (A and B) modified from those published by Lehman et al.
(1984). The individual cases are identified by numbers. Abbreviations are: OC, optic
chiasm; OT, optic tract; PVN, paraventricular nucleus; SCN, suprachiasmatic nucleus; 21,

zona incerta.

11

 

- . "PL-=6.-.“ ‘7.“

Figure 2: Photomicrograph of a coronal section (50 mm thick, pyronin Y stain) through the
hypothalamus of a hamster (No. 1) in which the knife cut (indicated by arrows) produced
bilateral damage just dorsal to the PVN. For abbreviations, see Figure 1.

12

 

 

34.

 

 

 

 

\s
s\

 

LGVG DAY SHORT DAY BILATERAL
OOVTRCI (INTRO. KC

 

 

Figure 3: Mean (iSEM) testicular weights of the short-day housed hamsters with bilateral
knife cuts (11 = 7) and sham-operated hamsters kept in short days (shaded bar; n = 9) or

long days (open bar; n = 6).

13

unilateral or undetectable also showed testicular regression (testicular weights were 0.45,
0.38, and 0.64 g, respectively) although they were not statistically compared with the
other groups.

Microscopic examination of the brains revealed that HRP/TMB reaction product was
present in the SCN of all eight hamsters that received injections of HRP. The distribution
of reaction product was similar to that reported in intact hamsters after intraocular injection
of HRP (Pickard and Silverman, 1981), indicating that the knife cuts did not damage the
retinohypothalamic tract.

Each of the 6 hamsters with bilateral knife cuts and the 3 control hamsters for
which activity patterns were monitored showed normal nocturnal patterns of activity.
Figure 4 shows the activity record of a hamster (No. 2) in which the knife cut was ventral
to the PVN. Onset of activity lagged behind onset of darkness (i.e., negative phase angle).
Mean phase angle (i SEM) was -238.5 :1: 34.0 min for the four hamsters with bilateral
knife cuts (Nos. 2, 5, 10, 11) and -156.0 i 6.0 min for control hamsters. Phase angle did
not differ statistically between hamsters with knife cuts and control hamsters (11 = 2.5,

p = 0.228). One hamster in which the knife cut produced only unilateral damage to the
hypothalamus (No. 7) and the hamster in which damage from the cut could not be detected
(No. 12) showed phase angles of -306 min and -72 min, respectively.

Discussion

Knife cuts ventral to, through, or dorsal to the PVN prevented testicular regression in
hamsters housed in nonstimulatory photoperiod without disrupting their circadian activity
patterns. This effect of knife cuts ventral (Eskes and Rusak, 1985; Inouye and Turek,
1986) or dorsal (Inouye and Turek, 1986) to the PVN on the testicular response to short-
day photoperiod has been independently shown by other investigators. Also, in rats, knife
cuts placed ventral to the PVN but dorsal to the SCN fail to affect free-running circadian
rhythms of locomotor activity and drinking behavior (Brown and Nunez, 1986).

14

No.2

—0
_N
.5

Figure 4: Activity record of a hamster (No. 2) that received a bilateral knife cut ventral to
the PVN and did not show testicular regression after 13 weeks of exposure to short days.

The black bar at the top represents the dark portion of the light-dark cycle.

15

It should be noted that sampling of testicular condition at only one point in time (via
castration) does not permit evaluation of the possibility that the effect of the knife cuts was
to affect the rate of testicular regression rather than to prevent this response. Therefore, in
Experiment 2B (below) repeated sampling of testicular width of hamsters was used to test
this possibility.

The knife cut ventral to the PVN probably prevented the testicular response to short-
day photoperiod by interrupting neural input to the PVN from the SCN (Berk and
Finkelstein, 1981; Silverman, et al., 1981; Stephan, et al., 1981). Efficacy of cuts through
the PVN may have been due to interruption of SCN input to PVN cells dorsal to the cut,
disruption of communication between cells within the PVN, and/or interruption of
descending projections from the PVN to the thoracic spinal cord (Armstrong, et al., 1980;
Don Carlos and Finkelstein, 1982; Swanson and Kuypers, 1980b).

Interpretation of the results of cuts dorsal to the PVN is less clear and at least two
possible explanations exist. The SCN send fibers through and past the PVN to the
habenula and other areas (Stephan, et al., 1981) and projections from the PVN to the
habenula (Dafny, 1983) and the pineal (Buijs and Pevet, 1980; Guerillot, et al., 1982;
Korf and Wagner, 1980; Moller and Korf, 1983) have been described. Thus, the dorsal
cuts as well as the PVN lesions reported by other investigators (Bartness, et al., 1985;
Klein, et al., 1983; Lehman, et al., 1984; Pickard and Turek, 1983;) should have
interrupted these pathways from both the SCN and the PVN. Therefore, the loss of the
gonadal response to short-day photoperiod might have been due to disruption of either of
these pathways. Alternatively, the dorsal cuts may have blocked testicular regression by
severing descending projections from the PVN to the thoracic spinal cord. Experiments 2
and 3 were conducted in order to further explore the neural pathways that mediate
testicular regression in hamsters exposed to short days. Experiment 2 addressed the
question of whether knife cuts dorsal to the PVN abolish testicular regression by

interrupting paraventriculo-spinal projections. Experiment 3 was designed to evaluate the

l6

possibility that the effects of knife cuts and electrolytic lesions in or near the PVN are due

to interruption of fibers that pass through the area of the damage.

EXPERIMENT 2

The purpose of this experiment was twofold: first, to determine whether knife cuts
dorsal to the PVN interrupt axons that project from the PVN to the spinal cord, and
second; to replicate the results of Experiment 1 using repeated measurements of gonadal
status. When the experiment was initiated, the exact course of fibers that project from the
PVN to the spinal cord had not been described. Previous investigators had observed axons
(e.g., Armstrong, et al., 1980) and axon collaterals (Hatton, et al., 1985) that project
dorsally as well as those that project laterally from neurons of the PVN of rats, although
limitations of the techniques that were used prevented the visualization of the full length of
the longer of these fibers from the cell bodies of origin to the termination of the axons.
Since these dorsally-projectin g fibers from cells of the PVN had been reported, and since
projections from the PVN to the spinal cord were known to exist both in rats (Armstrong,
et al., 1980; Swanson and Kuypers, 1980b) and in hamsters (Don Carlos and Finkelstein,
1982), it seemed plausible that knife cuts dorsal to the PVN might prevent testicular
regression by severing fibers that project from the PVN to the spinal cord. Further
evidence that is consistent with this possibility was provided by Nance (1981), who
reported that an obliquely parasagittal (or "frontolateral") unilateral knife cut prevented
transport of bisbenzamide from the spinal cord to the PVN of rats. Nance interpreted this
result as being due to disruption of fibers that project laterally from the PVN and terminate
in the spinal cord. Careful inspection of the tracings of representative brain sections
presented by the author, however, reveals that while the lower portions of these cuts were
lateral to the PVN, the orientation of these cuts was such that the dorsalmost tips of the
cuts were dorsal to the PVN and lay at the midline of the animal. Thus, the cuts that were
effective in abolishing projections from the PVN to the spinal cord would have severed
fibers that project dorsally as well as those that project laterally from the PVN. Therefore,
given the knowledge that was available, it was not possible to determine whether the

dorsal knife cuts that prevented testicular regression in Experiment 1 had severed

l7

18

projections from the PVN to the spinal cord.

-.r 5 ° Do Li) Q. horn. o " A r 111.0 '-__r:. 5191 .- 0.! -. in .06. ”'11 '7
Initially, unilateral horizontal cuts dorsal to the PVN were employed to further
investigate the neural pathways destroyed by the cuts that abolish photoperiodic
responses. These cuts were aimed at stereotaxic coordinates identical to those used in
Experiment 1 and were followed by bilateral injections of HRP into the upper thoracic
spinal cord. The procedure for these injections is explained in detail in Part B. Sections of
the spinal cords were microscopically inspected to verify that the injections were bilateral.
In addition an area of the lateral hypothalamus served as a control area to ensure that
asymmetries in the distribution of labelled PVN neurons were not due to failure of the
injections of HRP. labelled neurons were counted in an area of the lateral hypothalamus
represented in Figure 5. In order to show the location of knife cuts and the distribution of
labelled neurons in the PVN, tracings were made of a series of sections through the PVN.
In one case, the cut was placed dorsal to the anterior and medial portions of the PVN and
passed through the dorsalmost part of the lateral portion (the lateral parvocellular division
of the nucleus, according to the terminology of Swanson and Kuypers [l980b]). Similar
to the knife cuts used by Nance (1981), this cut was in a position in which it should have
severed most of the axons that project laterally, as well as most of those that project
dorsally out of the PVN. As illustrated schematically in Figure 6, > 90% of the labelled
PVN cells in this hamster were located contralateral to the knife cut. Conversely, labelled
cells in the lateral hypothalamus were relatively evenly distributed between the two sides
of the brain. The results of the cell counts indicated that 10 labelled cells were found in 7
sections through the control area ipsilateral to the cut while 12 were found on the
contralateral side. These results indicate that the cut probably severed most of the axons
that project from the PVN to the spinal cord. In another hamster (No. 128), an
asymmetrical distribution of labelled PVN neurons was found. The cut in this hamster was
more dorsal and centered more anterior than that in No. 122 and did not produce direct

damage to the PVN. This cut was in a position such that it should have severed some of
1 9

20

Figure 5: Location of the area (indicated by hatching) of the lateral hypothalamus in which
counts of HRP labelled cells were made. The drawings (A-C, rostral to caudal) of the
hamster hypothalamus are a modification of those published by Malsbury (1977).
Abbreviations are: ARC, arcuate nucleus; F, fomix; OT, optic tract; VMN, ventromedial

nucleus.

 

 

 

22

Figure 6: Schematic representation of the distribution of HRP—labelled neurons in three
sections (at 240 um intervals, A-C are rostral to caudal) through the paraventricular
nucleus of a hamster (No. 122) that received a unilateral knife cut (indicated by arrows)
before bilateral injections of HRP into the spinal cord. Note the reduction in the number of
HRP-labelled cells ipsilateral to the cut. Abbreviations are: 0C, optic chiasm; OT, optic
tract; PVN, paraventricular nucleus; SCN suprachiasmatic nucleus; ZI, zona incerta; 3V,

third ventricle. Each dot represents one labelled cell.

23

 

24

the fibers that project dorsally out of the PVN but probably did not damage axons that
project laterally out of the nucleus. The location of the cut and the distribution of labelled
neurons in this hamster are shown in Figure 7. The number of labelled neurons in 5
sections through the control area of the lateral hypothalamus of hamster No. 128 were 35
(ipsilateral to the cut) and 37 (contralateral to the cut). The small reduction in the number
of labelled neurons in the PVN ipsilateral to the cut in this hamster indicates that some of
the axons that project from the PVN to the spinal cord may have been severed by the cut
but that many were left intact. In a third hamster, the cut was just dorsal to the PVN. As
shown in Figure 8, the distribution of labelled PVN neurons was much more symmetrical
in this hamster than in either of the two cases discussed above.

Although no strong conclusions can be drawn from these findings, these results, as
well as the finding that bilateral cuts that were placed dorsal to the PVN prevented
testicular regression (Experiment 1) are consistent with the idea that some paraventriculo-
spinal fibers leave the nucleus in a dorsal direction while others take a lateral course upon
leaving the PVN. In contrast to the results of dorsal cuts, bilateral parasagittal cuts placed
lateral to the PVN fail to prevent testicular regression in hamsters exposed to
nonstimulatory photoperiod (Larry Morin, personal communication). Thus, it appears that
fibers that project laterally from the PVN are not critical for the testicular response to short
days.

Results published after these initial findings were obtained have provided more direct
evidence that some paraventriculo-spinal fibers project dorsally out of the PVN. Following
injection of the plant lectin, PHA-L, into the PVN of rats, Luiten, et al. (1985) were able
to trace two bundles of labelled fibers from the PVN through the dorsolateral funiculus of
the spinal cord to the intermediolateral cell column at the thoracolumbar level of the spinal
cord. The two bundles were reported to be composed of approximately equal numbers of
fibers. Bundle number 1 takes a dorsal course upon leaving the PVN and travels through

the thalamus and mesencephalic periaqueductal grey. Bundle number 2 leaves the PVN in

25

Figure 7: Schematic representation of the distribution of HRP-labelled neurons in three
sections (at 240 um intervals, A-C are rostral to caudal) through the paraventricular
nucleus (PVN) of a hamster (No. 128) that received a unilateral knife cut (indicated by
arrows) before bilateral injections of HRP into the spinal cord. Note that the cut was more
dorsal and centered more anterior than that in hamster No. 122 (Figure 5). Note also that
the reduction in the number of HRP-labelled cells ipsilateral to the cut was less substantial

than that in hamster No. 122.

26

 

Figure 7

27

Figure 8: Schematic representation of the distribution of HRP-labelled neurons in three
sections (at 240 um intervals; A-C, rostral to caudal) through the paraventricular nucleus
(PVN) of a hamster (No. 123) that received a unilateral knife cut (indicated by arrows)
before bilateral injections of HRP into the spinal cord. Note that the cut was dorsal to the
PVN and failed to produce a reduction in the number of HRP-labelled cells in the PVN

ipsilateral to the cut.

28

 

Figure 8

29

a lateral and caudal direction turning caudally at the boundry between the internal capsule
and the medial forebrain bundle. At the level of the pens, fibers from the two bundles
merge and continue coursing caudally. Preliminary results obtained using PHA-L indicate
that paraventriculo—spinal fibers in the hamster may follow a course similar to that reported
in the rat (Youngstrom and Nunez, 1986). Provided that similar pathways exist in the
hamster, it would appear that the knife cut through the dorsal PVN prevented transport of
HRP into the PVN by severing both bundle 1 and bundle 2. More dorsally placed knife
cuts would have spared most or all of the fibers in bundle 2 and hence, failed to have as
dramatic an effect on the retrograde transport of the enzyme into the PVN. Furthermore,
some of the fibers of bundle 1 may arise as collateral branches of axons of bundle 2. If a
sizeable proportion of these fibers are collateral branches of the same parent axon, then
transection of only one of these two bundles would not be likely to produce a dramatic
reduction in the number of neurons labelled after injections into the spinal cord. While
Luiten, et al. did not address the issue of axon collaterals, Hatton, et al. (1985) have
observed axons of magnocellular PVN neurons that give rise to collateral branches. While
most of these branches terminated within the hypothalamus, some of them apparently left
the plane of the sections without showing a terminal thickening and thus, may have
projected to extrahypothalamic sites. Therefore, while the possibility that some of the
fibers in bundle l and bundle 2 may arise as collateral branches of the same parent axon is
speculative, evidence of branching of axons of PVN neurons has been reported.

Since the results of Luiten, et al. (1985) provided evidence that the dorsal cuts in
Experiment 1 probably interrupted bundle 1 and that similarly placed unilateral cuts would
not sever all of the fibers that project from the PVN to the spinal cord, the strategy
involving unilateral cuts and spinal injections of HRP was abandoned. Work in progress
(Y oungstrom and Nunez) using PHA-L, should provide direct evidence for the course of
fibers that project from the PVN to the spinal cord in hamsters. The use of PHA-L allows

direct visualization of axons and, therefore, is a more powerful technique for elucidating

30

the course of fiber tracts than placing knife cuts in their path. Instead, Part B of the present
experiment was focused on replicating the results of Experiment 1 and determining
whether it is possible to prevent testicular regression by severing some but not all of the

PVN projections to the spinal cord (i.e., bundle 1 but not bundle 2).

-._r o' m. 'Hi ._0- '1'. .3 0‘ Hr 1.3.1' 'I' 1‘ 01017 0 in .0.

Since testicular condition was measured only once in Experiment 1, it was not
possible to determine whether the knife cuts had prevented testicular regression or affected
the rate of testicular regression. It is remotely possible that the animals with confirmed
bilateral cuts showed an accelerated testicular regression followed by a rapid spontaneous
gonadal recrudescence. It is also possible that the effect of the cuts was to slow the rate or
delay the onset of testicular regression. Although 8-12 weeks of exposure to short days or
blinding is generally considered sufficient to induce complete testicular regression in
hamsters, there may be conditions under which regression is delayed in some animals. In
one report (Eskes and Zucker, 1978) 3 out of 9 hamsters that were housed in activity
wheels showed little or no testicular regression 13 weeks after blinding. Two of these 3
animals showed signs of testicular regression 17 weeks after blinding. Thus, since it is
possible to delay gonadal regression, it was important to replicate Experiment 1 using
repeated measurements of testicular condition in order to evaluate the time-course of
gonadal regression or lack thereof. Moreover, the use of repeated measurements on
individual animals facilitated assessment of the possibility that partial knife cuts might
cause partial effects such as delayed gonadal regression.

Bilateral knife cuts were made dorsal to, or through the PVN of hamsters before the
animals were exposed to short days. Gonadal status of the animals was monitored via
biweekly measurement of testicular width. At the end of the experiment, horseradish
peroxidase was injected bilaterally into the spinal cord of hamsters in which the cuts had
prevented testicular regression to determine whether any projections from the PVN to the

spinal cord still existed in these animals.

31

Methods

Subjects and housing conditions were similar to those reported in Experiment 1.
Procedure

Twelve hamsters were given bilateral knife cuts aimed dorsal to the PVN as described
in Experiment 1. Ten animals served as sham-operated controls. All of these animals were
housed in long-day photoperiod before and for a week after surgery. All of the hamsters
with knife cuts and five of the sham-operated control hamsters were then moved to a
short-day photoperiod while five sham-operated hamsters served as long-day photoperiod
control animals. After four weeks of housing in short days, the animals were anesthetized
with methoxyflurane (Metofane; Pittman-Moore, Inc.) and testicular length and width
were recorded to the nearest 0.1 m using hand-held calipers. This procedure was
repeated biweekly for 12 weeks (i.e. until the hamsters had been exposed to the
nonstimulatory photoperiod for 16 weeks). Three hamsters with knife cuts were sacrificed
after 12 weeks of exposure to short days. At the completion of data collection, the 9
remaining hamsters with knife cuts and the 10 sham-operated hamsters were anesthetized
and castrations were performed as previously described. The 9 animals with knife cuts
were allowed to recover fi'om this surgery for one week before they were transferred to
activity wheel cages and their activity patterns monitored as in Experiment 1. One hamster
failed to drink after being placed in the activity wheel and was sacrificed before the
completion of data collection. As a result, wheel-running activity data were recorded for
only 8 of the hamsters. Phase angles of their activity patterns were estimated from ll-day
segments of the activity records as described in Experiment 1.

Bilateral injections of HRP into the spinal cord were performed in 5 animals under
Equithesin anesthesia. An incision was made through the skin on the dorsal surface of the
animal. With the aid of a dissecting microscope, adipose, muscular, and connective tissue
were then dissected and retracted to expose the vertebral column. A lO—ul Hamilton
syringe with a 32-gauge needle was used to inject 2 ul of a 30% solution of HRP (Sigma

32

33

Type VI) into the spinal cord between the seventh cervical (C7) and first thoracic (Tl)
vertebrae. The injections were bilateral and were made midway between the midline and
the lateral edge of the spinal cord. Injections were made over a l-min period and the needle
left in place for approximately 1 min after the injection. Following the injections of HRP,
the hamsters were allowed to survive 48 hr before perfusion-fixation as described in
Experiment 1. Hamsters that did not receive injections of HRP were perfused with 0.9%
saline followed by 10% formaldehyde.
Histology

For animals that had received injections of HRP, frozen oblique coronal sections of
the brains (3040 um thick) and transverse sections of the spinal cords (3080 um thick)
were reacted with TMB as in Experiment 1. For each animal, alternate sections of the brain
and spinal cord were counterstained with pyronin Y. Other series of brain sections were
counterstained with cresylecht violet while the other series of spinal sections were either
counterstained with cresylecht violet or dehydrated in ethanol and coverslipped unstained.
One hamster died before HRP injections were scheduled. Brains of the hamsters that did
not receive HRP injections were stored in 10% formalin with 30% sucrose until
sectioning. Alternate sections were mounted on slides and stained with cresylecht violet.
Microscopic examination was used to verify the extent and location of knife cuts and the
presence or absence of labelled cells. A drawing tube attached to the microscope was used
to trace selected sections. From these tracings, schematic diagrams were made to illustrate
the extent and location of the knife cuts at the anterior-posterior level of the PVN and,
where applicable, the distribution of PVN neurons containing HRP reaction product.
Statistical comparisons were made using two-tailed Mann-Whitney 11 tests, corrected for
multiple comparisons, when necessary, as in Experiment 1.

Results

In 5 animals, the knife cuts failed to produce the intended bilateral damage to the

hypothalamus throughout the rostro-caudal level of the PVN. In 3 of these 5 hamsters

34

(Nos. 102, 107, 108) the cuts produced asymmetrical damage that was mostly or entirely
anterior to the PVN. In one hamster (No. 109), the cut produced damage dorsal to the
anterior part (approximately 1/3) but not the more posterior portions of the PVN. No
evidence of the knife cut was seen in the last hamster (No. 116). In all of these cases, the
cuts failed to prevent testicular regression. Mean testicular width of these hamsters was
7.14 i 0.64 mm after 12 weeks of exposure to short days. For the 4 of these animals that
were allowed to survive for 16 weeks of exposure to short days (Nos. 102, 107, 108,
116), mean testicular weight at that time was 0.964 :1: 0.1199 g. No clear evidence for
partial effects of misplaced or asymmetrical cuts was found, however, one of these
animals showed evidence of spontaneous testicular recrudescence during the experiment.
For hamster No. 108, testicular width reached a minimum of 5.7 mm at week 10 of
exposure to short days and grew to 9.1 mm by week 16 as shown in Figure 9.
Microscopic evaluation of the brains indicated that in 7 animals, the cuts produced
bilateral damage to the hypothalamus through most or all of the rostro—caudal level of the
PVN. In one of these cases (No. 103), the knife cut produced a bilateral mechanical lesion
of the PVN. In the other six hamsters that received bilateral damage to the hypothalamus,
the cuts were centered anterior to the cuts made in Experiment 1 (and anterior to the
intended location). In all six of these cases (see Figure 10) the cuts did, however, produce
bilateral damage through (No. 114) or dorsal to (Nos. 104, 105, 111, 115, 118) the PVN
through most or all of its anterior-posterior extent. All 7 of these animals with damage in
or dorsal to the PVN maintained large testes (i.e., 210.6 mm in width), as shown in
Figure 11. Mean testicular width after 12 weeks of exposure to short days was
12.26 :1: 0.278 mm. Two of these hamsters were sacrificed after 12 weeks of exposure to
short days. Mean paired testicular weight for the remaining 5 animals (Nos. 103, 105,
111, 114, and 115) was 3.776 :1: 0.4048 g after 16 weeks of exposure to the
nonstimulatory photoperiod. In addition, the five sham-operated control hamsters housed

in long-day photoperiod all maintained adult gonadal size, showing a mean testicular width

 

 

 

 

 

 

 

14
2/
12 -- 6’ A»
°‘—--o
E10 ..
f I
5 8 " '0' K0102
E 6
§ -°- K0107 ~41 A
:1 6 " -I- K0108
Q
:7, 4 ,_ 13- K0109
Lu
1— *- KC116
21- 4' SHORT DAY
CONTROL (11:5)
0 : t i i i : .
2 4 6 8 1o 12 14 16

Figure 9: Testicular widths of individual animals in which the knife cuts failed to produce
the intended bilateral damage dorsal to or through the PVN (solid lines) and mean

testicular width of control hamsters (n = 5) exposed to short days (broken line).

36

SUN

 

 

 

UT

 

 

 

 

Figure 10: Schematic representation of the location of knife cuts that produced bilateral
damage to the hypothalamus in Experiment 2. The cuts are drawn to show their location
and extent in sections (A or B) through the PVN although in most cases the cuts were
centered anterior to the nucleus. In one hamster (No. 103, not shown, the cuts produced a

bilateral mechanical lesion of the PVN). Abbreviations are as in Figure 1.

37

 

 

 

 

 

 

 

14 -
E12 "
.5,
E10 --
9
3
g 8 “I N
a 6 "
m a
'2 41 " bilateral KC's (nus?)
E '0' long day control (n-S)
2
2 JL ‘9' short day control (n-5)
o t . . . . : .
2 4 6 8 1o 12_ 14 16

Figure l 1: Mean testicular width for the 7 hamsters with bilateral knife cuts (KC) through
or dorsal to the PVN, for 5 sham-operated control hamsters that were housed in short-day
photoperiod, and 5 sham-operated hamsters that were housed in long-day photoperiod.
Data for one hamster (No. 103) in which the knife cut produced a mechanical lesion of the
PVN are included with data for the other hamsters that received knife cuts. Two of the

hamsters that received knife cuts were sacrificed early, hence, the last two data points

(weeks 14 and 16) do not include data for them.

38

of 12.7 :1: 0.094 mm at week 12 and a mean paired testicular weight of 4.728 31: 0.0670 g
at week 16 of exposure of short-day housed animals to that photoperiod. In contrast,
control hamsters that were housed in short days all showed complete testicular regression,
exhibiting a mean testicular width of 7.22 :1: 1.034 mm at week 12 and a mean paired
testicular weight of 0.751 t 0.0755 g at week 16 of exposure to the nonstimulatory
photoperiod. As illustrated in Figure 11, the average testicular width of control hamsters
housed in short days for 12 weeks was statistically significantly different both from that of
control hamsters housed in long days (11 = 0, p = 0.008) and from that of the 7 hamsters
with extensive bilateral damage through or dorsal to the PVN (1.1 = 1.5, p < 0.01), while
the latter two groups did not differ statistically from one another (1.1 = 9.5, p = 0.202).
Similarly, at week 16 the average paired testicular weight of the control hamsters housed
in short days was statistically significantly different from that of control animals housed in
long days and from that of the 5 hamsters with bilateral knife cuts through or dorsal to the
PVN (11 = 0, p = 0.008, for each comparison) while the difference between the latter two
groups did not reach statistical significance (1.1 = 5, p = 0.15).

All 8 of the hamsters for which locomotor activity was recorded exhibited normal
nocturnal patterns of activity. In all of these cases, the circadian activity pattern showed
normal entrainment to the light-dark cycle and a negative phase angle. Mean phase angle
was -315 i 33.27 min for hamsters with bilateral knife cuts that prevented testicular
regression (Nos. 105, 111, 114, and 115) and -283.5 i 11.33 min for hamsters in which
the knife cuts were asymmetrical (Nos. 102, 107, 108) or undetectable (No. 116) and
failed to prevent testicular regression. The difference in average phase angle between these
two groups did not reach statistical significance (L1 = 4, p = 0.342).

Bilateral injections of HRP were made in 5 of the 7 hamsters that received knife cuts
that prevented testicular regression. In one of these animals, labelled cells were not found
in either the PVN or in the lateral hypothalamus, thus, it appears that there was no
transport of the enzyme into the hypothalamus. In 4 of these animals, however, HRP

39

injections and TMB reactions resulted in unequivocal labelling of cells in the PVN. In
three hamsters in which the knife cuts were dorsal to the PVN, a number of labelled
neurons were found in the PVN. The locations of the knife cuts and distributions of
labelled neurons in two animals, representative of these cases, are presented in Figures 12
(No. 115) and 13 (No. 118). A darkfield photomicrograph showing the knife cut dorsal
to, and labelled neurons in the PVN of one of these hamsters (No. 118) is presented in
Figure 14. The knife cut in the fifth hamster that received injections of HRP (No. 114)
passed through the PVN. A few labelled neurons were found in the PVN of this animal
(data not shown) both ventral and dorsal to the cut.

I? .

Knife cuts through or dorsal to the PVN again prevented testicular regression in
hamsters exposed to short days but failed to affect their entrained circadian activity
rhythms, thus confirming the results of Experiment 1. There was no evidence that
testicular regression had begun in any of these animals by week 16 of exposure to short
days. These results indicate that the bilateral cuts disrupted the neural pathways that
mediate effects of photoperiod on the reproductive system but did not affect the neural
mechanisms responsible for the generation and entrainment of circadian activity rhythms.
Furthermore, these activity patterns all showed relatively large negative phase angles.
Elliot (1976) reported that intact hamsters housed in long days (11.5-18 hrs of light/ 24
hrs) show activity patterns in which the active phase of the rhythm is compressed and the
phase angle of the rhythm varies with the length of the photoperiod. With increasing
length of the dark portion of the light-dark cycle, the active phase of the activity rhythm
becomes decompressed, reaching an asymptote at a photoperiod of about 11.5 hr. Further
increases in the length of the dark portion of the light-dark cycle have little effect on the
length of the active phase of the rhythm and, moreover, the onset of activity always falls
about 12 hrs before the termination of darkness. Thus, in hamsters housed under a light-
dark cycle, circadian rhythms are entrained by the onset of light. All 8 of the animals for

40

Figure 12: Schematic representation of the distribution of HRP-labelled neurons in three
sections (at 240 um intervals; A-C, rostral to caudal) through the paraventricular nucleus
(PVN) of a hamster (No. 115) that received a bilateral knife cut (indicated by arrows) that
prevented testicular regression. This animal received bilateral injections of HRP into the
spinal cord and was sacrificed 48 hr later. Note that the cut was dorsal to the PVN and
failed to prevent labelling of cells in the PVN.

41

 

Figure 12

42

Figure 13: Schematic representation of the distribution of HRP-labelled neurons in three
sections (at 240 um intervals; A-C, rostral to caudal) through the paraventricular nucleus
(PVN) of a hamster (No. 118) that had received a bilateral knife cut (indicated by arrows)
that prevented testicular regression before receiving bilateral injections of HRP into the
spinal cord. Note that the cut was dorsal to the PVN and failed to prevent labelling of cells
in the PVN.

43

1111

 

 

Figure 13

44

 

Figure 14: Darkfield photomicrograph of a section (30 pm thick, counterstained with
pyronin Y) through the hypothalamus of a hamster (No. 118) in which a bilateral knife cut
(indicated by large arrows) just dorsal to the PVN prevented testicular regression but did
not prevent labelling of PVN neurons (indicated by small arrows) after injections of HRP

into the spinal cord.

45

which activity patterns were monitored in the present experiment showed phase angles
within the range reported by Elliot (1976) for hamsters housed in a photoperiod of 6 hrs
light:18 hrs dark. Thus, while the reproductive system of hamsters with cuts through or
dorsal to the PVN responded as if the animals were exposed to long days, the animals
responded behaviorally as if they were exposed to short days.

In 3 hamsters, cuts dorsal to the PVN prevented testicular regression but failed to
prevent retrograde transport of HRP from the spinal cord to the PVN. This indicates that
some paraventriculo-spinal fibers were spared by the cuts and suggests that destruction of
all paraventriculo-spinal fibers is not necessary to prevent the gonadal response to short-
day photoperiod. These cuts should have interrupted all of the fibers of bundle 1 but
probably did not sever all of the fibers of bundle 2.

In contrast to the results seen in hamsters with bilateral cuts, exposure of animals
with asymmetrical or misplaced knife cuts to short days resulted in testicular widths S 7.0
mm in all cases. Furthermore, the length of time required for animals with these cuts to
attain testicular widths of S 7.0 mm was within the range of the time required for control
hamsters to reach the same criterion. Thus, there was no evidence of partial effects of
these cuts.

In contrast to the results of dorsal cuts, parasagittal cuts aimed lateral to the PVN fail
to block testicular regression in hamsters housed in nonstimulatory photoperiod (Larry
Morin, personal communication). Taken together, these results suggest that bundle l
mediates responses of the pineal and the gonads to photoperiod while bundle 2 does not.
Bundle 2 appears likely to be involved in other autonomically mediated responses. For
example, parasagittal knife cuts placed in the path of bundle 2 result in hyperphagia and
obesity in rats (Gold, et al., 1977). However, it appears that bundle 1 is also involved in
the regulation of energy balance since knife cuts placed in a coronal plane through the
central grey of the thalamus or midbrain and transecting the path of bundle 1 (Weiss and
Leibowitz, 1985) abolish the feeding response that is normally elicited in satiated, intact

46

rats by injection of norepinephrine or clonidine into the PVN (Leibowitz, 1978; Weiss and
Leibowitz, 1985). Further evidence implicating bundle 2 in the regulation of energy
balance is the finding that electrical stimulation centered at a site lateral to the anterior
PVN, in or near bundle 2 evokes an initial decrease followed by a nearly 1° C rise in the
temperature in interscapular brown adipose tissue (BAT [Brown, 1986]). Both in rats

(T ulp, et al., 1982) and hamsters (Bartness and Wade, 1984; Wade, 1982, 1983), dietary
composition influences metabolic efficiency (i.e. the proportion of ingested calories that
are diverted to energy storage depots) and thermogenesis. Furthermore, photoperiod also
influences metabolic efficiency and thermogenesis in hamsters (Bartness and Wade, 1984;
Campbell, et al., 1983; Wade, 1983). These effects are mediated in part by brown adipose
tissue (Campbell, et al., 1983; Tulp, et al., 1982). While descending neural pathways to
BAT have not yet been fully mapped, recent evidence suggests that the PVN may be
involved in the regulation of thermogenesis (Bartness, et al., 1985).

It is interesting to note that in one case a cut dorsal to the anterior 1/3 of the PVN
failed to prevent testicular regression. The anterior-posterior placement of the cut in this
animal is reminiscent of that of a partial electrolytic lesion reported by Lehman, et al.
(1984). The lesion was reported to have "spared most of the mid- and caudal PVN" while
destroying the more anterior portion of the nucleus. Blinding of the animal resulted in
substantial but incomplete testicular regression. Taken together, these results hint that the
anterior part of the PVN is not critical for testicular regression in hamsters exposed to

nonstimulatory photoperiod.

EXPERIMENT 3

In addition to their projections to the PVN, the SCN send fibers dorsally through
and by the PVN to the habenula, periventricular thalamus and the mesencephalic central
grey (Stephan, et al., 1981). Therefore, it is possible that electrolytic lesions of the PVN
and knife cuts through and dorsal to the nucleus block testicular regression by interrupting
fibers that do not terminate in the PVN. To test whether cells of the PVN are, in fact,
involved in the gonadal response to short-day photoperiod, the present set of experiments
employed intrahypothalamic injections of N-methyl-aspartate (NMA), a neurotoxin which
destroys cell bodies while apparently sparing fibers that pass through the injection site
(Hastings, et al., 1985; Olney and Price, 1983).

47

Mcthxlasrattatfl

Injections of NMA, a structural analog of the putative neurotransmitter aspartate, into
the lateral hypothalamus in the region of the medial forebrain bundle destroy lateral
hypothalamic neurons without affecting dopamine or serotonin content of the basal ganglia
(Hastings, et al., 1985b). Thus, the toxin appears to selectively destroy neurons without
affecting axons that pass through the site of the injection. While magnocellular neurons of
the PVN appear to be insensitive to the toxic effects of NMA, evidence suggests that
parvocellular neurons of the PVN are destroyed by the toxin (Olney and Price, 1983).
Furthermore, a recent report indicates that kainic acid, a structural analog of glutamate,
destroys parvocellular but spares magnocellular neurons of the PVN (Zhang and Ciriello,
1985).

Hastings, et al. (1985a) reported that injections of NMA into the anterior
hypothalamus, in an area dorsolateral to the SCN, prevented testicular regression and
suspension of vaginal cyclicity in hamsters exposed to nonstimulatory photoperiod. These
authors reported that the injections that blocked the effects of exposure to short days had
caused a loss of cells in the anterior hypothalamic area (AHA) dorsolateral to the SCN
without affecting either the SCN or the PVN. The fact that the animals in the study by
Hastings, et al. exhibited normal circadian rhythms of locomotor activity indicates that the
SCN were functionally intact. Ftn'thermore, other investigators have reported that neurons
of the SCN are resistant to the toxic effects of kainic acid (Peterson and Moore, 1980).
However, since the figures presented by Hastings, et al. indicate that the site of the
injections that blocked gonadal regression included at least part of the PVN, it is suprising
that they reported that the injections did not affect neurons within the PVN.

In order to determine whether neurons of the PVN are sensitive to the toxic effects of

NMA, the present experiment employed unilateral intrahypothalamic injections of NMA.

48

49

These intrahypothalamic injections were followed by bilateral injections of HRP into the
spinal cord to determine whether the toxin kills cells that project to the spinal cord.
Methods

Male hamsters were housed in long days as described in Experiment 1. Hamsters
(n = 5) were anesthetized with Equithesin (4.5 to 5 ml/kg body weight) and placed in a
stereotaxic apparatus. A l-ul Hamilton syringe was then used to inject N-methyl-D,L-
aspartate (NMA; Sigma Chemical Co., St. Louis, MO) dissolved in artificial cerebrospinal
fluid (ACSF) on one side of the brain. The constituents of the ACSF were 130 mM N aCl,
25 mM NaHCOg, SmM NazI-IPO4, 30 mM KCl, 8 mM MgClz, and 13 mM CaC12
(Dohanich and Clemens, 1981). The injection volume (0.5 ul), concentration of NMA
(0.12 M), and pH of the solution (7.0) were chosen so as to replicate parameters used by
Hastings, et al. (1985). For two animals in which the injections were aimed at the PVN,
stereotaxic coordinates were 0.8 mm posterior to the bregma, 6.0 mm ventral to the
superior sagittal sinus and 0.5 mm lateral to the midline with the top of the incisor bar
2.0 mm below the ear bars. For the rest of the animals (n = 3) the injections were aimed at
the AHA, and the anterior-posterior stereotaxic coordinate was changed to 0.2 mm anterior
to the bregma. To reduce backflow of the solution up the needle tract, injections were
made over a 2-min period and the needle was left in place for 5 min after the injection.

Animals were allowed to survive for at least 6 days before injections of HRP were
made into the spinal cord as described in Experiment ZB. The hamsters were allowed to
survive for 48 hr after the injections of HRP and then sacrificed using an overdose of
Equithesin. Perfusions and reaction and staining of the tissue were performed as in
Experiments 1 and 2. Thickness of the sections was 30 um for the brains and 40-80 um
for the spinal cords. Microscopic evaluation of the tissue was performed to determine the
locations of the NMA injections and to visualize the HRP reaction product in the brain and
spinal cord. For quantification of loss of cells that project to the spinal cord, labelled

neurons in the PVN on each side of the brain were counted using the pyronin-Y stained

50

sections through the nucleus (8 sections/brain). Cell counts were made by two
investigators and labelled cells from 8 sections taken at random were independently
counted by both investigators to test the reliability of the method. Although the variance in
the number of HRP-labelled cells differed between the two sides of the brain, the
Student's t test for independent samples (2-tailed probabilities) was used for statistical
comparisons. For comparisons involving groups of equal sizes, the outcome of the
Student's t test is not seriously affected by differences in variance between groups (e.g.;
see Boneau, 1960). Thus the Student's 1; test, which is more powerful than the Mann-
Whitney U test for detecting differences between groups, was used.
Results

The NMA injections were centered near the PVN in 4 hamsters. In 3 of these animals,
the injections were centered dorsolateral to the PVN while in the fourth hamster the
injection was centered ventrolateral to the PVN and dorsolateral to the SCN. The center of
the injection sites (i.e., termination of the needle tracks) in these hamsters are represented
schematically in Figure 15. In a fifth hamster, the injection site was near the midline,
rostral to the preoptic area and just dorsal to the diagonal band of Broca.

Microscopic examination of the brains of the hamsters in which injections were
centered near the PVN indicated a loss of neurons in the posterior PVN ipsilateral to the
injection site. Neurons in the anterior PVN of these animals, however, were spared.
Moreover, the loss of HRP/TMB labelled neurons was particularly dramatic (see Figure
16). As shown in Figure 17, counts of labelled neurons revealed a statistically significant
loss of labelled cells ipsilateral to the NMA injections (1 = 3.07, df = 6, p < 0.05). In these
4 hamsters, the PVN contralateral to the injection site contained an average of
89.65 :1: 3.18% of the total number of labelled cells. For the sections that were evaluated
by two investigators, the independent estimates were in good agreement with
91.68 :1: 2.79% and 89.92 i 2.64% of the total number of labelled cells found on the

uninjected side of the brain. The difference between the two estimates did not reach

51

Figure 15: Schematic representation of the locations of the centers of the unilateral
injections of NMA in 4 hamsters that received bilateral injections of HRP into the spinal
cord after the NMA injections. The drawings of the hamster hypothalamus (A-C are
rostral to caudal) are modifications of those published by Malsbury (1977). Abbreviations
are: ARC, arcuate nucleus; F, fomix; OT, optic tract; PV, paraventricular nucleus; SC,

suprachiasmatic nucleus; SO, supraoptic nucleus.

53

Figure 16: Photomicrographs of a series of sections (30 um thick, counterstained with
pyronin Y) at 180 um intervals (A—D, rostral to caudal) through the PVN of a hamster
(No. 380) that received a unilateral injection of NMA just dorsolateral to the PVN prior to
bilateral injections of HRP into the spinal cord. Note that in the most anterior section (A)
there does not appear to be any loss of unlabelled cells on the injected (left) side; however,
the number of both HRP-labelled, and unlabelled cells on the injected side is substantially

reduced in the more posterior sections (B-D).

54

       

       
   

! ~\ V .4 _ , "
1' 1233541. fic- :
' I mi} . 111311" if; :5- .

H-.:;~..;.' 12:-1 3. 49?}. ~.'» 1 '3'

   

4 ‘I I‘ f
.. _’_

z
r, .
'r

it:

j"6.;'6\' ‘,3
r id. '§. .

2" 13.3% "

" -'T H!

.. ,. {a
‘A, _ r-‘.o" .
.4" :.-*-,-~.-,,:\-~ 1;?

.';. 5

J?“

a
v
v
‘1
‘Q

  
   

     

 

1
’h‘" 1' TM ' (L 5‘3 I.» "
“Wkly x 3 "I ¢~-->
be .. ,1, ‘ . _ - ‘ , “
"‘3 1‘ "t. '. S; 1 o- ., 1,...
‘ ~ ‘31,. I? ‘1, - .,'._~.‘L~§:

t)
I*.
. u
‘ 3'11
~
.0
O
‘V
o
-9
:3:
. I‘ l!
M‘-

4
If“

. am;
oh"

 

55

600 r

01
O
O

 

A
O
O

 

200 ‘

100:

NUMBER OF LABELLED NEURONS
8
O

 

 

 

 

 

1
/////W

CONT RALATERAL IPSILATERAL

 

 

Figure 17: Mean (iSEM) number of HRP-labelled neurons in the PVN ipsilateral and
contralateral to the site of injection of NMA. The centers of the sites of the NMA injections

in the four hamsters were represented in Figure 15.

56

statistical significance (1 = 0.49, df = 14, p > 0.05). In contrast, the hamster in which the
NMA injection was centered near the midline showed a more symmetrical pattern of
labelled neurons in the PVN (47.18 vs 52.82%).

E . .

These results support earlier reports that intrahypothalamic injections of the aspartate
analog, NMA (Olney and Price, 1983), and the glutamate analog, kainic acid (Zhang and
Ciriello, 1985), destroy parvocellular but spare magnocellular neurons of the PVN.
Furthermore, since unilateral injections of the toxin resulted in a statistically significant
reduction in the number of PVN neurons containing HRP reaction product after injections
of the enzyme into the spinal cord, the present results extend previous findings in
demonstrating that neurons that project from the PVN to the spinal cord are among the
cells that are sensitive to the toxic effects of NMA. Since all of the injection sites and
needle tracks were outside of the PVN, it is not likely that the loss of cells in the PVN was
due to direct mechanical damage produced by the needle. The use of unilateral injections of
NMA permitted utilization of the uninjected side of each brain as a control in evaluating the
damage produced by the NMA. This procedure circumvents some of the problems
associated with comparisons of number of HRP labelled cells across animals (e.g.,
differential transport or enzymatic activity between cases). Similar results have recently
been reported independently by other investigators. Although the amount of destruction
was not quantified, Hastings and Herbert (1986) reported that injections of NMA
destroyed parvocellular PVN neurons that project to the spinal cord (as determined by
injections of true blue into the spinal cord), but not axons (as observed by the Holmes
silver stain method) or magnocellular neurons within the PVN.

In summary, the present results indicate that intrahypothalamic injection of NMA
should serve as an effective tool in studying the function of neurons that project from the

PVN to the spinal cord.

’91,: a o - ”n: h i .. ' o 2', .. n ' .l.._ s. 2' o J!J.o.!010f
1' {o ‘ o " x -1'10 "run-Dwain -._ no. (‘01-...f‘fi 7
Since the results of Experiment 3A indicated that NMA does kill nemons of the PVN

and particularly neurons that make paraventriculo-spinal connections, a further experiment
was executed in which bilateral injections of NMA were used to test whether neurons of
the PVN are involved in the testicular response to exposure to short days. In some of the
animals, injections were aimed at the PVN in order to maximize neuronal destruction in
that nucleus. In additional animals, the injections were aimed at the AHA dorsolateral to
the SCN in an attempt to replicate the results of Hastings, et al (1985a). In addition to the
use of testicular width and weight as measures of gonadal function, ability of some of the
animals to inseminate female hamsters was tested.

Methods

Subjects and housing conditions were the same as reported in Experiment 1.
5mm

Under Equithesin anesthesia, hamsters received one of three types of stereotaxically-
placed injections. Bilateral injections of NMA (0.5 ul of 0.12 M NMA dissolved in
ACSF, pH 7 .0) were aimed at either the AHA or the PVN, while control injections of
(0.5 pl of) the ACSF vehicle were aimed at the PVN. To prevent accidental
contamination, separate syringes were used for injections of the toxin and the ACSF
vehicle. The injection procedure was as described in Experiment 3A. Stereotaxic
coordinates for injections aimed at the PVN were 0.8 mm posterior to the bregma,
:1: 0.5 mm lateral to the midline, and 6.1 mm ventral to the superior sagittal sinus, For
injections aimed at the AHA, stereotaxic coordinates were 0.2 mm anterior to the bregma,
0.5 mm lateral to the midline, and 6.0 mm ventral to the sinus. In all cases the top of

incisor bar was positioned 2.0 mm below ear bar zero.

57

58

Rmedute

Hamsters that received bilateral injections of NMA (n = 27) or ACSF (n = 13) were
returned to long-day photoperiod and allowed to recover from surgery for approximately 1
week. These animals were then transferred to short-day photoperiod. Five hamsters that
did not receive injections were individually housed in long-day photoperiod to serve as
control animals for that photoperiod. Begining 4 weeks after the hamsters were moved to
short-day photoperiod and their testicular width and length recorded to the nearest 0.1 mm
biweekly as reported in experiment 2B. For testicular measurements, the animals were
injected with Ketamine HCl (V etalar, Parke-Davis; 10 mg/ 100 g body weight).Surgeries
were performed on two batches of hamsters staggered 6 weeks apart. Some of the animals
from the second batch were used in an experiment (described below) to determine whether
hamsters that have been experimentally induced to maintain adult gonadal size during
exposure to short days will show sexual behavior when presented with the opportunity to
do so. Because this experiment required testing of the animals after they should have
shown testicular regression, the length of time between transfer of the hamsters to short
days and sacrifice of the animals varied between the two batches. After 14 weeks of
exposure to short days, 10 hamsters that had received injections of NMA (8 that had
exhibited testicular regression and 2 that had failed to do so), and 4 of the control hamsters
housed in short days were anesthetized with Equithesin and castrated. Paired testicular
weights for these animals were obtained to the nearest 0.1 mg. These hamsters were
sacrificed by an overdose of the anesthetic after castration. Hamsters that had received
injections of NMA were perfused as described below. Of the 31 hamsters that were
allowed to survive beyond week 14 of the experiment, 17 were sacrificed at week 19. The
17 hamsters that were sacrificed at that time included 2 that exhibited complete testicular
regression and 5 that showed no, or only partial testicular regression after injections of
NMA, 5 control hamsters that were housed in short days and 5 control hamsters that were

housed in long days. These animals were anesthetized and their testes and seminal vesicles

59

removed. Paired weights of each organ were recorded to the nearest 0.1 mg. These
hamsters were then given an overdose of the anesthetic and animals that had received
injections of NMA were perfused as described below.

The remaining 14 hamsters (10 that received NMA injections and 4 that received
control injections) had all shown complete testicular regression. These hamsters were
allowed to survive until week 25 of the experiment. From week 19 to week 25, the width
of the right testis of each of these hamsters was recorded biweekly as described above in
order to determine whether these animals were capable of exhibiting testicular
recrudescence. One of the hamsters that had received NMA injections died between week
25 and week 28 of the experiment. The brain of this animal was removed and stored in
10% formaldehyde. At week 28, testicular and seminal vesicle weights for the 13
remaining animals were obtained, the animals sacrificed, and perfusions performed on
hamsters that had received injections of NMA, as described below.

Histology

After castrations had been performed, the hamsters that had received injections of
NMA were perfused transcardially with 0.9% saline followed by 10% formaldehyde.
Brains were removed and stored in 10% formaldehyde with 30% sucrose. Frozen
sections, 40 um thick, were cut through the preoptic area and hypothalamus. Alternate
sections were mounted on slides, stained with cresyl violet and coverslipped. In order to
facilitate evaluation of the tissue without knowledge of the gonadal response of individual
animals, identification numbers of the animals were reassigned at this time by an
investigator who did not participate in the histological analysis.Microscopic examination
(at magnifications of 25 to 400 X) of the brains was performed in order to determine the
extent and location of the damage produced by the NMA injections. In performing this
analysis, the amount of neuronal loss in the PVN was noted. For each animal, tracings of
the location of the center of the injection sites (i.e., termination of the needle track) were

made via the use Of a drawing tube attached to the microscope. These tracings were then

60

used to identify the locations of the centers of the injection sites on a standard set of
drawings of the hamster hypothalamus that were modifications of the drawings published
by Malsbury (1977). These hamsters were then divided into three groups based on the site
and extent of neuronal damage. Hamsters in the first ("complete" damage) group showed
extensive loss of neurons from the PVN while animals in the second (partial damage)
group showed substantial but partial loss neurons fiom the PVN and animals in the third
(no damage) group showed either minimal or no loss Of neurons from the PVN. After this
histological evaluation was performed, reproductive responses of the animals that had
received injections of NMA were compared with those of control hamsters. Statistical
comparisons of testicular widths and testicular and seminal vesicle weights between
hamsters that had shown extensive damage to the PVN and the two control groups, and
between the two groups of animals that were allowed to show testicular recrudescence
were made using Mann-Whitney 11 tests corrected for multiple comparisons, when
necessary, as described in Experiment 1. The time of sacrifice of hamsters from each of
the 3 histologicallydefined groups and each of the 2 control groups are presented in Table
1.

E l . l I .

During weeks 16 and 17 of exposure of the hamsters to short days, reproductive
competence of 5 hamsters in which injections of NMA had produced complete or partial
insensitivity to short days was verified via behavioral testing of the ability of these animals
to deliver sperm. Under Equithesin anesthesia, female hamsters were ovariectomized
through a single abdominal incision. Estrogen replacement was delivered via Silastic
capsules implanted subcutaneously on the ventral surface of the animal. Each capsule
consisted of a segment of Silastic medical grade tubing (Dow Corning) 3.175 mm 0. D.,
1.575 mm 1. D. filled with crystalline estradiol benzoate (Steraloids, Inc.) and sealed on
each end with wooden plugs covered with Silastic sealer. The distance between plugs was

3-5 mm. Sexual behavior of the 5 male control hamsters exposed to long days and the 5

61

Table 1: Number of hamsters sacrificed at each of three times and number of hamsters that
were used in the sexual behavior tests.

 

 

Group Wed Number Used in
(Type of Damage) Number Week 14 Week 19 Week 29 Behavioral Tests
Long Day Control 5 0 5 0 5

Short Day Control 13 4 5 4 0
NMA-Injected 27 10 7 10 5
("Complet6" (5) (1) (4) (0) (4)
Damage)

(Partial Damage) (6) (2) (3) (1) (1)

(N0 Damage) (16) (7) (0) (9) (0)

 

62

male hamsters that maintained the largest testes after injections of NMA were tested in
glass aquaria containing a small amount of bedding. A dim red incandescent bulb, turned
on before the onset of darkness and left on overnight, provided illumination (and a
romantic atmosphere) during the behavioral observations. Testing sessions for the control
hamsters began at the begining of the dark portion of the light-dark cycle while testing of
the animals that received injections of NMA began 8 hr after the begining of the dark
portion of the light dark cycle, in an attempt to minimize differences due to testing the two
groups during different phases of their circadian activity patterns. Since Elliot (1976)
found that hamsters housed in photoperiods of 18 and 6 hr of light begin to show wheel
running activity 2 hr before and 6 hr after the onset of darkness, respectively, the testing
sessions for each group began 2 hr after the average onset of activity as measured by
wheel running. In order to induce sexual receptivity, female hamsters were given
subcutaneous injections of 0.5 mg of progesterone (Sigma) in 0.05 ml of sesame oil, 4 hr
before the begining of the testing session. For each test, a male was placed in an aquarium
(51 x 26 x 30 cm) and allowed to habituate to the testing conditions for 5 min before
introduction of the female. Behavior of the animals was Observed for 20 min and the
aquaria were washed and bedding replaced between tests. In the first test (week 16),
sexual behavior was not quantified but exhibition of mounting behavior was noted for
each animal. One week later, sexual behavior of the animals was again tested and recorded
orally on a Panasonic F35 cassette tape recorder. For quantification, behavioral measures
were scored as latency to show first mount and first intromission after introduction of the
female, and number of mounts during which intromission did not occur and number of
intromissions exhibited during the 20—min testing session. Statistical comparisons between
the two groups were performed using Student's t test for independent samples (two-tailed
probabilities). The rationale for the use of this test was as discussed in Experiment 3A.
Since all of the hamsters failed to ejaculate during the 20-min testing sessions, ability of

the animals that received NMA injections to ejaculate was tested by pairing these animals

63

with sexually receptive females overnight on the day before they were sacrificed. Five of
the ovariectomized females received subcutaneous injections of progesterone (0.5 mg in
0.05 ml of sesame oil) 4 hr before termination of the dark portion of the light-dark cycle.
Each of these animals was placed in a cage with a male shortly before the end of the light
portion of the light-dark cycle. The following morning vaginal smears were Obtained from
the females by insertion of a cotton swab, moistened with distilled water, into the vagina.
Fluid from the swab was deposited onto a microscope slide. As a test of the ability of
these hamsters to exhibit spermatogenesis, a small amount of seminal fluid was removed
and placed onto microscope slides following castration of the animals. Vaginal and
testicular smears were microscopically examined for evidence of spermatozoa.
Results

As shown in Figure 18, all 13 of the control hamsters that were housed in short days
showed complete testicular regression by week 14 of the experiment while the 5 control
hamsters that were housed in long days maintained large testes for the duration of the
experiment. Mean testicular widths for these two groups were 5.89 :t 0.17 and
12.9 :1: 0.43 mm, respectively. This difference was statistically significant (11 = 0,
p, = 0.00138).
EEE “11“.“. 1 . IE I II . l B

The sites of the injections of NMA typically showed a zone of dense gliosis along the
path of the needle surrounded by a region of less dense gliosis and loss of neurons. In 5
hamsters, the NMA injections caused a substantial loss of neurons from the PVN,
bilaterally. As illustrated schematically in Figure 19, the injections were centered just
dorsal or dorsolateral to the PVN in 3 (Nos. 1, 2, 3) of these hamsters and within the
boundaries of the PVN in 2 additional animals (Nos. 23, 28). In all 5 of these hamsters,
neurons in the anterior PVN, for the most part, survived the injections. In contrast, the
posterior and lateral portions of the PVN were completely, or almost completely destroyed
by the NMA injections, thus the bilateral pattern of damage in these animals‘was similar to

64

 

 

 

 

 

 

 

 

 

14 --
9 b Tc_____ 9
. t:- . —_ I
.5,
E10 1
Q
3
m 8 «t-
5 \
3 .
«1-
: 6 O
11"}
g 4 1‘ 'I' “complete" damage (0.5)
g '0- Iong day control (n-5)
2 " "' short day control (n-13)
0 : . : : 4 :
2 4 6 8 10 12 14
WEEKS

Figure 18: Mean testicular width of 13 control hamsters that were housed in short days
after receiving injections of ACSF, 5 unoperated control hamsters that were housed in
long days, and 5 hamsters that were housed in short days after receiving injections of

NMA that caused "complete" damage to the PVN.

65

\s y
\ 0 QM
A‘

 

B

\
a
V 1.1

Figure 19: Schematic representation of the locations of the centers of the injections of

/

 

 

NMA that produced "complete" bilateral destruction of the PVN. The pairs of circled
numbers represent the bilateral injection sites in individual hamsters. The drawings of the
hamster hypothalamus (A-B, rostral to caudal) are modifications of those published by
Malsbury (1977). Abbreviations are: ARC, arcuate nucleus; F, fomix; IC, internal
capsule; OT, optic tract; MAH, medial anterior hypothalamus; MFB, medial forebrain
bundle; MT, mammillo-thalamic tract; PV, paraventricular nucleus; RT, reticular thalamic

nucleus; V, ventral thalamic nucleus.

66

the unilateral pattern of damage in the hamsters that received unilateral injections just
dorsolateral to the PVN in Experiment 3A. In 4 cases a small proportion of cells in the
posterior and periventricular parts of the nucleus survived after the injections. In all 5 of
these cases there was minor damage to the area of the dorsal hypothalamus surrounding
the PVN, as well as the ventral thalamus including the zona incerta. Photomicrographs of
a series of sections through the area of the PVN of a hamster representative of these cases
are presented in Figure 20. All 5 of these hamsters maintained adult testicular size for the
duration of the experiment Mean testicular widths of these animals are presented in Figure
18. After 14 weeks of exposure to short days, the average testicular width of these animals
(12.3 i 0.29 mm) was statistically different from that of the control hamsters that were
housed in short days (1.1 = 0, p = 0.00138) but not from that of control hamsters that were
housed in long days (L1 = 8, p = 0.420). One of these hamsters (No. 1) and 4 of the
control hamsters that were housed in short days were sacrificed after week 14 of the
experiment. Testicular weight of the hamster that sustained "complete" damage to the PVN
was 2.54 g while the mean testicular weight of the control hamsters was 0.445 :1: 0.042 g.
Five control hamsters that were housed in each of the two photoperiods aird 4 hamsters
that had sustained "complete" damage to the PVN were sacrificed at week 19 of the
experiment. Mean testicular weights for these groups were 2.679 t 0.168 g for hamsters
that sustained "complete" damage to the PVN, 3.403 :1: 0.137 g for the hamsters that were
housed in long days and 0.837 t 0.124 g for the control hamsters that were housed in
short days. Testicular weight of the control hamsters housed in short days was statistically
significantly different both from that of control hamsters housed in long days (11 = 0,

p = 0.008) and from that of hamsters that showed "complete" damage to the PVN Q1 = 0,
p = 0.016). The difference in testicular weight between the animals that sustained
"complete" damage to the PVN and hamsters that were housed in long days did not reach

statistical significance (U = 2, p = 0.064). Mean seminal vesicle weights for hamsters that

67

Figure 20: Photomicrographs of a series of 40 pm thick sections (160 um intervals, cresyl
violet stain) through the hypothalamus of a hamster (No. 3) representative of those cases
in which the loss of neurons of the PVN, following bilateral injections of NMA, was
judged to be "complete" damage. Notice that while PVN neurons are visible in the more
anterior sections (A and B), there was a substantial loss of neurons from the posterior

PVN (C and D). The arrows in D indicate the gliosis surrounding the needle tracks.

 

Figure 20

69

sustained "complete" damage to the PVN, and for control hamsters housed in long, and
short days were 1.020 t 0.339 , 1.412 :t 0.198, and 0.125 t 0.023 g, respectively. On
this measure, control hamsters that were housed in short days differed statistically both
from control hamsters housed in long days (11 = 0, p = 0.008) and from hamsters that
showed "complete" damage to the PVN (U = 0, p = 0.016), while the difference between
animals that sustained "complete" damage to the PVN and hamsters that were housed in
long days did not reach statistical significance (U = 7, p = 0.556). Mean testicular and
seminal vesicle weights of these groups of hamsters are presented in Table 2.

In 6 hamsters that received NMA injections, a loss of neurons which was
independently judged by each of 2 investigators to be partial damage to the PVN was
found. As shown in Figure 21, the injections of NMA were centered dorsal to the PVN in
5 of these hamsters and dorsal to the SCN in the other animal. Testicular widths of
individual hamsters in this group are shown in Figure 22 while testicular and seminal
vesicle weights are given in Table 2. In one hamster (No. 15), the loss of PVN neurons
on one side of the brain was comparable to that in hamsters that sustained "complete"
damage, while the contralateral PVN showed a less extensive loss of neurons. This
hamster maintained large testes for the duration of the experiment, showing a testicular
width of 12.1 mm (see Figure 22) and a testicular weight of 2.330 g when sacrificed at
week 14. The hamster (No. 4) in which the NMA injections were centered dorsal to the
SCN showed substantial but partial damage to the PVN bilaterally and exhibited partial
testicular regression. Testicular width of this hamster at week 14 was 9.2 mm (see Figure
22). The other hamsters that showed partial damage to the PVN included one (No. 8) that
showed "complete" damage to the PVN on one side and no damage to the contralateral
PVN, and 3 hamsters (Nos. 7, 14, 21) that showed partial damage to the PVN on each
side. As shown in Figure 22, the injections of NMA failed to prevent testicular regression

in all of these animals.

Table 2: Testicular and seminal vesicle weights*.

70

 

 

Week of Testicular Seminal Vesicle
Group Sacrifice Weight (g) Weight (g)
Short Day 14 0445:0042 .....
Control
"Complete" 14 2.540 -----
Damage
Partial 14 2.330 """
Damage 0.6104
NO Damage 14 0.58711:0.057 -----
Short Day 19 0.837i0.124 0. 125120.023
Control
Long Day 19 3.403i0.137 1.412i0.198
Control
"Complete" 19 2.679i0.168 1.020i0.339
Damage
Partial 19 0.6280 0.1375
Damage 1.4676 0.4888
Short Day 29 3.695:l:0. 104 l.643:1:0.166
Control
No Damage 29 3.389:1:0.266 1.180:1:0. 199
Partial 29 3.885 1.907
Damage

 

*Weights are reported as group meaniSEM, where feasible. For hamsters that received

partial damage to the PVN and the one hamster that was sacrificed at week 14 after

receiving "complete" damage to the PVN, data for individual animals are reported.

72

 

 

Figure 21: Locations of the centers of the injections of NMA that produced partial damage
to the PVN. The drawings of the hamster hypothalamus (A-D, rostral to caudal) are
modifications of those published by Malsbury (1977). Abbreviations are: F, fomix; IC,
internal capsule; OT, optic tract; MAH, medial anterior hypothalamus; MFB, medial
forebrain bundle; PV, paraventricular nucleus; RT, reticular thalamic nucleus; SC,

suprachiasmatic nucleus; 80, supraoptic nucleus; V, ventral thalamic nucleus.

 

 

TESTICULAR WIDTH (mm)
m

 

 

 

 

 

5 ._
4 ..
1‘" 21
2 1r
)0 short day
0 fi' i 3 t : 4
2 4 6 8 10 12 14

Figure 22: Testicular widths of individual hamsters in which the injections of NMA

(presented in Figure 21) produced partial damage to the PVN.

74

In 7 of the hamsters that were sacrificed at week 14 of the experiment, the injections of
NMA did not cause damage to the PVN. The centers of the injection sites in these animals
are represented in Figure 23. As shown in that figure, the placement of the injections in
these hamsters was quite variable, ranging from just dorsal to the anterior commissure at
the level of its decussation across the midline, to the central region of the thalamus at the
anterior-posterior level of the PVN. In each case, gliosis and loss of neurons were seen in
the area surrounding the injection site. All of these hamsters showed complete testicular
regression. Mean testicular width of these animals at week 14 was 5.71 i 0.45 mm. Mean
testicular weight for this group (see Table 2) was 0.587 i 0.057 g.

Of the hamsters that were sacrificed at week 19 of the experiment, 2 animals (Nos. 14
and 21) showed partial damage to the PVN (discussed above) and complete testicular
regression. Testicular widths (Figure 22) of these animals at week 14 were 5.6 mm (No.
14) and 6.7 mm (No. 21). At the time of sacrifice, one of these hamsters showed evidence
of having begun to exhibit spontaneous testicular recrudescence. Testicular weights of
these animals were 0.6280 and 1.4676 g, while the corresponding seminal vesicle weights
were 0.1375 and 0.4888 g (T able 2). This observation led to a major change in procedure.
The remaining animals (10 that had received NMA injections and 4 control animals) had all
shown complete testicular regression and had been scheduled to be sacrificed at week 19.
These animals were allowed to survive an additional 9 weeks and their testicular widths
recorded biweekly. While this manipulation further complicated the experimental design, it
pemritted a physiological assay of the ability of these animals to secrete LH and FSH and,
by implication LHRH. In one of these hamsters (No. 8), the injections caused partial
damage to the PVN (discussed above). In the other 8 hamsters that were sacrificed at week
28 and the hamster that spontaneously died after week 25, the injections did not cause
damage to the PVN. The locations of the centers of the NMA injection sites that did not
damage the PVN, represented in Figure 24, were similar to the locations of the injections

 

 

 

Figure 23: Locations of the centers of the injections of NMA that did not damage the PVN
or prevent testicular regression in hamsters that were exposed to short days for 14 weeks.
The drawings of the hamster hypothalamus (A-E, rostral to caudal) are modifications of
those published by Malsbury (1977). Abbreviations are: AC, anterior commissure; F,
fomix; IC, internal capsule; OC, optic chiasm; OT, optic tract; MAH, medial anterior
hypothalamus; MFB, medial forebrain bundle; MPO, medial preoptic area; PV,
paraventricular nucleus; RT, reticular thalamic nucleus; SC, suprachiasmatic nucleus; SM,

stria medullaris; SO, supraoptic nucleus; V, ventral thalamic nucleus.

 

 

 

 

Figure 24: Locations of the centers of the injections of NMA that did not damage the PVN
of hamsters that were exposed to short days for 25-28 weeks. The drawings of the
hamster hypothalamus (A-E, rostral to caudal) are modifications of those published by
Malsbury (1977). Abbreviations are: AC, anterior commissure; F, fomix; IC, internal
capsule; OC, optic chiasm; OT, optic tract; MAH, medial anterior hypothalamus; MFB,
medial forebrain bundle; MPO, medial preoptic area; PV, paraventricular nucleus; RT,
reticular thalamic nucleus; SC, suprachiasmatic nucleus; SM, stria medullaris; SO,

supraoptic nucleus; V, ventral thalamic nucleus.

79

that did not damage the PVN of hamsters that were sacrificed at week 14. In Figure 25 a
photomicrograph of a section just caudal to the center of the injection site of one of these
hamsters is presented. As shown in that figure, the injection caused substantial gliosis and
loss of neurons in the central part of the thalamus, but did not damage the PVN. As
illustrated in Figure 26, all 9 of these hamsters as well as the control hamsters that were
sacrificed at week 28 showed complete testicular regression by week 14 and complete
testicular recrudescence by week 25. Mean testicular widths of control hamsters and
hamsters that received NMA injections that did not damage the PVN were 6.10 :1: 0.21 and
6.34 :1: 0.37 mm at week 14 and 12.03 :1: 0.47 and12.38 :1: 0.18 mm at week 25. The
difference between these two groups was not statistically significant at either of these two
times (L1 = 15, p = 0.64552, at week l4;L1 = 18, p = 1.0, at week 25). Furthermore,
neither the difference in testicular weight (3.389 1 0.266 vs 3.695 :1: 0.104 g, 11 = 11,
p = 0.28014) nor the difference in seminal vesicle weight (1.18 i 0.199 vs
1.643 :t 0.166 g, 11 = 6, p = 0.06432) between the 8 hamsters that survived until week 28
after receiving NMA injections that did not damage the PVN and the control group reached
statistical significance (see Table 2). The hamster that had shown partial damage to the
PVN also exhibited complete testicular regression and recrudescence, having shown
testicular widths of 6.2 mm at week 14 and 13.1 mm at week 25, and testicular and
seminal vesicle weights of 3.885 g and 1.907 g at week 25.
Behavioral Tests

During the first testing session, the 5 hamsters that received NMA injections (Nos. 2,
3, 4, 23, and 28) and the 5 control hamsters all showed mounting behavior. During the
second testing session, all of the males again showed mounting behavior. The hamster
(No. 4) that had exhibited partial testicular regression failed to show intromission while
the other 9 hamsters all displayed this aspect of sexual behavior. None of the hamsters
ejaculated during the 20-min testing session. As illustrated in Figures 27 and 28, however,

the animals that received NMA injections were less sexually competent than control

80

 

Figure 25: Photomicrograph of a section (40 pm thick, cresyl violet stain) through the
brain of a hamster in which the NMA injections were centered in the thalamus
approximately 750 um dorsal to the PVN. The gliosis and loss of neurons is typical of
cases in which the NMA injections were centered in the thalamus. Note that the damage
did not extend to the PVN. The open arrow indicates the zone of dense gliosis

surrounding the center of the NMA injection site.

_s
A
l

MEAN TESTICULAR WIDTH (mm)

 

2

—L

N
L
U

1o-L

 

 

"' nma (n=9)

‘0' acsl (n=4)

 

 

I
U

345

II

6 7

81

./'7:j
\-\.___..-_./...—o

k j I I

I I I I
r I U T U I I'

I I l I I I I
T I U I V U V U

8 910111213141516171819202122232425
WEEKS

Figure 26: Testicular widths of the 9 hamsters for which the injection sites were presented

in Figure 24 and for 4 control hamsters that were exposed to short days for 25- 28 weeks.

82

hamsters. While the average number of mounts did not differ statistically (j; = 0.28, df = 8,
p > 0.05) between the animals that received NMA injections (5.8 i: 1.9) and control
hamsters (5.2 :1: 0.04), hamsters that received injections of NMA showed a longer latency
to mount (173.0 1 31.1 vs 93.0 :1: 15.3 sec, 1 = 2.31, df = 8, p < 0.05), fewer
intromissions (27.8 :1: 1.7 vs 42.4 i 2.6, t = 4.41, df = 8, p < 0.01), and a longer latency
to intromit (272.5 :1: 53.8 vs 132.3 i 9.8 sec, 1 = 2.89, df = 8, n < 0.05) than control
animals. Although none of the hamsters showed the full complement of male sexual
behavior during the sessions in which they were observed, 4 out of the 5 that had received
NMA injections were capable of delivering sperm. Microscopic examination of the vaginal
smears revealed that intact spermatozoa, as well as detached tails were present in the
vaginae of the females that were paired with these males. No evidence of spermatazoa was
found in the vaginal smear obtained from the female that was paired with the hamster (No.
4) that had shown partial testicular regression. Intact spermatozoa were found in the
testicular fluid Obtained from each of the 5 hamsters that received NMA injections.

D. .

In 6 animals, injections of NMA produced a substantial loss of neurons from the
posterior PVN as judged independently by two investigators. As reported in Experiment
3A, and by other investigators using NMA (Olney and Price, 1983) or kainic acid (Zhang
and Ciriello, 1985), the magnocellular neurons of the anterior portion of the PVN were not
destroyed by injections of the toxin. In all six of these cases, the injections of NMA
prevented testicular regression when the hamsters were exposed to short days. In 4 of
these animals, two measures of testicular function, in addition to weight and size, were
used to confirm that the gonads of these animals had remained active. Since seminal
vesicle weight is dependent on circulating concentrations of gonadal steroids (Ellis and
Turek, 1979), the large seminal vesicles of these animals indicates that their testes were
secreting androgens. Moreover, the presence of spermatozoa in the vaginal smears

obtained from the 4 female hamsters that were paired overnight with these animals

83

 

 

NUVlBEROF MOINTS
A

 

 

 

 

 

51 y/
T? fl

Figure 27: Mean number of mounts (without intromissions) and latency to mount

 

to
01
j

 

c-L
0|
0

d
O
O

 

 

LATENCY TO MOUNT (sec)
0'1
0

 

 

 

 

 

 

 

exhibited by hamsters that received injections of NMA (n = 5) that caused either complete
(11 = 4) or partial (n =1) damage to the PVN, and unoperated control hamsters that were
housed in long days. While the number of mounts did not differ statistically between

groups, the difference in latency to mount was statistically significant (p < 0.05).

84

 

 

 

 

 

 

 

 

 

50‘
8
E 30. ,
S 201 //
§ 10:
z . %
A 4001
i.
5 300‘
23
5
E 200'
g
.9 i
5 1001
,5.
S C

 

 

 

 

 

NMA OWTRC].

Figure 28: Mean number of intromissions and latency to intromit exhibited by hamsters
that received injections of NMA (n = 5) that caused either complete (n = 4) or partial

(n =1) damage to the PVN, and unoperated control hamsters that were housed in long
days. Statistically significant differences between groups were obtained both for number

of intromissions (p < 0.01) and for latency to intromit (p < 0.05).

85

indicates that they were able to produce sperm and inseminate females. While limitations
of the techniques that were used prohibited assessment of the viability of the spermatozoa,
the results suggest that reproductive competence of the animals was maintained.

Some of the hamsters received injections that were aimed at the area of the anterior
hypothalamus which Hastings, et al. (1985) reported that NMA-induced damage of
prevented reproductive regression in male and female hamsters exposed to short days.
Since the majority of the injections in the present experiment were centered outside of this
area, it was not possible to compare these results with those of Hastings, et al. In one
notable case (No. 4), however, the injections were centered dorsolateral to the SCN
approximately midway between that nucleus and the PVN and within the area of neuronal
loss reported by Hastings, et al. (see their figure 1). The injections in this animal resulted
in a loss of neurons from the anterior hypothalamus similar to that reported by Hastings, et
al. and a partial loss of neurons from the PVN. In agreement with previous reports
(Peterson and Moore, 1980), no evidence of damage was seen in the SCN of this animal.
Exposure of this hamster to short days resulted in partial testicular regression.
Furthermore, there was no evidence of spermatozoa in the vaginal smear obtained fiom the
female hamster with which this animal was paired overnight. Microscopic examination of
the testicular fluid obtained from this animal after castration revealed that spermatozoa
were present. Thus, although this hamster may have been able to produce sperm, he failed
to deliver it. Of the injections that produced damage to the PVN, this one was centered the
farthest from the PVN. Since the glial scar could be traced from the center of the injection
site to the PVN, it is likely that the damage to the PVN was due to backflow of the toxin
along the path of the needle.

While injections of NMA that produced "complete" damage to the PVN were centered
either within the PVN or S 400 um dorsal (not corrected for shrinkage of the tissue during
histological processing) to it, injections that were centered 2 520 um dorsal to the PVN
and injections that were centered anterior to the PVN failed to produce damage to that

86

nucleus. Injections that were dorsal to the PVN and S 500 um from that nucleus produced
effects that ranged from "complete", to partial, to no damage to the PVN. In one animal
(discussed above), injections dorsal to the SCN, and ventral and anterior to the PVN
produced partial damage to the PVN. In one additional hamster, the injection on one side
of the brain was dorsal to the SCN while the contralateral injection was in the dorsal aspect
of the preoptic area. The injections in this hamster produced no or only very little loss of
cells from the PVN.

Injections that did not produce damage within the PVN failed to prevent testicular
regression. Furthermore, 9 of these hamsters were exposed to short days for 25 weeks
and all of them showed complete spontaneous testicular recrudescence. The ability of these
animals to show complete testicular recrudescence implies that the portion of the
hypothalamo-hypophysial axis responsible for the secretion of LH was not destroyed by
the injections of NMA. While the location of the LHRH neurons responsible for the
secretion of LHRH into the hypothalamo-hypophysial portal system is not specifically
known, the cells that are currently thought to be the most likely candidates lie in a ventral
position through the preoptic area and hypothalamus. Thus, that in each of the animals in
which NMA injections did not damage the PVN the injections were centered dorsal to
PVN (n = 10) in the thalamus, or in the dorsal aspect of the septum or preoptic area
(n = 5), and (with the exception of one animal in which the injection on one side was
dorsolateral to the SCN while the injection on the contralateral side was in the dorsal
aspect of the preoptic area) in no case was a loss of cells within the ventral hypothalamus
evident, support the argument that the injections in these animals probably did not cause
significant damage to the LHRH neurons that regulate release of gonadotropins fiom the
pituitary. Research in progress (Brown, Badura, and Nunez) using hamsters exposed to
long days after receiving injections of NMA should add further support to the argument
that NMA-induced neuronal damage similar to that reported here does not induce gonadal

regression.

87

As mentioned previously, partial destruction of the posterior PVN prevented testicular
regression in one hamster and partially prevented regression in another animal. In one
additional hamster in which partial destruction of the PVN was found, testicular regression
was delayed. Testicular width of this hamster was 10.5 mm at week 12 and 7.5 mm at
week 14. The 3 Other hamsters that received partial damage to the PVN all showed
complete testicular regression within the range of time required for control hamsters to do
so. Thus, the effects of partial destruction of paraventriculo-spinal projections on the
testicular response to short days were quite variable, ranging from no effect to delay or
partial prevention, to complete prevention of the response. Evidence for effects of partial
destruction of the PVN and its connections with the spinal cord on the testicular response
to short days has been reported previously. As mentioned above, Lehman, et al. (1984)
reported that a partial lesion of the PVN, which destroyed the anterior PVN but spared
most of the mid and caudal part of that nucleus, partially prevented testicular regression.
Moreover, in a study of the effects of knife cuts ventral and dorsal to the PVN on testicular
regression in hamsters housed in short days, Inouye and Turek (1986) reported variable
effects of their sham-surgical procedure. Their knife cuts were made with a non-retractable
wire knife that extended 1.5 mm perpendicularly from a stainless steel cannula. In order to
avoid damaging the superior sagittal sinus, the authors lowered the knife into the brain at a
point lateral to the midline, pushed the knife to the midline, and lowered the blade to a
point ventral to the PVN. This procedure was reported to result in "variable damage" to
areas dorsal to the PVN (e.g., hippocampus, thalamus, and habenular area) and either
prevented or partially prevented testicular regression in 7 out of 12 animals subjected to
this procedure. Moreover, that bilateral cuts (present study) placed dorsal to the PVN in an
area in which they should have completely severed bundle 1 but not bundle 2, prevented
testicular regression but not labeng of PVN neurons after injections of HRP into the
spinal cord (Experiments 1 and 2) supports the argument that testicular regression can be

prevented by severing only a portion of the paraventriculo—spinal projections. Since the

88

pattern of secretion of melatonin, gonadotropins, and other endocrine hormones in these
animals is not known, it is not possible to determine the mechanism by which misplaced
or partial lesions partially prevent testicular regression. It is possible, however that these
lesions affect the release of gonadotropins.

Although injections of NMA in the dose used in this investigation have been reported
not to damage monoaminergic axons that pass through the lateral hypothalamus (Hastings
et al., 1985b), toxic effects of NMA on axons of SCN nemons have not been directly
tested. However, converging evidence from experiments on the function of the SCN
indirectly suggests that these axons are probably insensitive to the toxicity of NMA. Knife
cuts that create an "island" of tissue containing the SCN and sever all of the efferent
projections of that nucleus are equivalent to SCN lesions (Stephan and Zucker, 1974) in
that they abolish circadian rhythms of locomotor activity, sleep, brain temperature, and
drinking behavior (Stephan and Nunez, 1977). Knife cuts that partially isolate the SCN,
severing fibers that project dorsally, laterally, and caudally, but not rostrally from the SCN
abolish the nocturnal pattern of water intake of rats housed under a light-dark cycle (Nunez
and Stephan, 1977). Thus, disruption of a large proportion of the axons that project from
the SCN causes a disruption of behavioral circadian rhythms. Further evidence regarding
disruption of axons of SCN neurons comes from a comparison of the effects of
intrahypothalamic injections of NMA with effects of injections of colchicine. Intracerebral
injections of colchicine temporarily disrupt microtubule—dependent axoplasmic transport
(Dahlstrom, 1968). Injections of colchicine have been used to produce temporary
behavioral deficits (Avrith and Morgenson, 1978; Willis and Smith, 1983). Bilateral
injections of colchicine (1, 2, or 4 ug in 0.2 ul of vehicle) just dorsolateral to the SCN
temporarily disrupt circadian rhythms of drinking behavior in blinded rats for up to 12
days, after which the rhythms return (Brown and Nunez, unpublished observation).
Injection of NMA (0.5 1110f 0.15 M) at the same site had no effect on the drinking
rhythm Furthermore, the NMA injection sites that were reported by Hastings, et al.

89

(1985a) to prevent gonadal regression in hamsters included, and appear to have
surrounded (see their figure 2), the SCN but did not affect the circadian activity patterns of
the animals. Since manipulations that are known to affect axons (i.e., knife cuts and
injections of colchicine) disrupt behavioral circadian rhythms while injections of NMA at
sites that include the SCN do not, these results indirectly suggest that axons of SCN
neurons are not damaged by injections of NMA. Furthermore, Hastings and Herbert
(1986) recently reported seeing intact axons (stained via the Holmes silver method) both
within the PVN and coursing dorsally through the site of damage produced by injections
of NMA into the PVN. Thus, it appears that the injections of NMA that produced
substantial damage to the PVN and prevented testicular regression in the present
experiment probably did so without destroying axons that have been reported to travel
from the SCN through or near the PVN to the habenula and other structures (Stephan, et
al., 1981). Therefore, the effects of electrolytic lesions or knife cuts in or near the PVN on
the testicular response to short days are probably not due to incidental damage to these
fibers of passage. Work in progress (Brown, Badura, and Nunez), using knife cuts aimed
just ventral to the habenula, may provide further evidence to support the argument that
projections from the SCN and the PVN to the habenula are not involved in testicular
regression in hamsters exposed of short days.

A recent report has provided evidence in support of the argument that knife cuts
(Experiments 1 and 2), injections of NMA (present experiment), and electrolytic lesions in
or near the PVN (Bartness, et al., 1985; Lehman, et al., 1984; Pickard and Turek, 1983)
prevent testicular regression by interrupting the neural pathway fiom the SCN to the pineal
and not the pathway by which pineal melatonin influences release of gonadotrOpins.
Injections of NMA that completely destroyed the parvocellular portions of the PVN
prevented the nighttime rise in pineal melatonin content (Hastings and Herbert, 1986).
Similarly, electrolytic lesions of the PVN also prevent the nocturnal increase in melatonin

content of the pineal (Klein, et al., 1983; Lehman, et al., 1984). Further research (e.g.,

90

testing of hamsters with electrolytic lesions of the PVN for sensitivity to systemic
injections of melatonin) is needed to confirm that the insensitivity to short days seen in
hamsters with damage to the PVN is not due to insensitivity to melatonin. Similarly,
further research (e.g., subhabenular knife cuts or electrolytic lesions of the habenula) is
needed to confirm that effects of damage to the PVN are not due to disruption of
projections from the PVN to the habenula (Conrad and Pfaff, 1976) or pineal (Buijs and
Pevet, 1980; Guerillot, et al., 1982; Korf and Wagner, 1980; Moller and Korf, 1983).
However, the most parsimonious explanation for the effects of electrolytic lesions,
injections of NMA or knife cuts in or near the PVN on the testicular response to short days
is that all of these manipulations disrupt the multisynaptic neural pathway from the SCN to
the pineal gland.

GENERAL DISCUSSION

These results are consistent with, and support the model of photic and circadian
regulation of the autonomic nemal and endocrine systems wherein the SCN act as a
circadian oscillator to regulate the timing of various functions. Mammalian circadian
rhythms persist in constant environmental conditions and are generated by one or more
circadian oscillators (Rusak and Zucker, 1979). The 1i ght—dark cycle acts as a zeitgeber
(time- giver) to entrain circadian rhythms. Ablation of the SCN abolishes several circadian
rhythms including adrenal corticosterone content (Moore and Eichler, 1972), pineal
serotonin N-acetyltransferase activity (Moore and Klein, 1974), and locomotor activity
and drinking behavior (Stephan and Zucker, 1972), as well as seasonal and ovulatory
cycles (Rusak and Zucker, 1979).

Through their axonal projections to other neural structures, the SCN produce seasonal
and circadian (and other) cycles of various end points. Thus, while axons that project
dorsally from the SCN to the PVN and other structures are not critical for the generation or
entrainment of behavioral circadian rhythms (Experiment 1; Brown and Nunez, 1986;
Eskes and Rusak, 1985), they appear to be necessary for photoperiod—dependent
reproductive cycles (Experiment 1; Eskes and Rusak, 1985; Inouye and Turek, 1986) and
photic and circadian regulation of pineal melatonin secretion (Lehman, et al., 1984; Klein,
et al., 1983; Moore and Klein, 1974). Similarly, results of experiments using
retrochiasmatic knife cuts (in a coronal plane caudal to the SCN) indicate that axons that
project caudally from the SCN are not critical for the generation of circadian rhythms of
water intake (Nunez and Stephan, 1977), feeding (Nishio, et al., 1979), or locomotor
activity (Nunez and Casati, 197 9); but appear to be necessary for such endocrine cycles as
daily surges of prolactin during pseudopreganancy (Freeman, et al., 1974), estrous cycles
(Nunez and Casati, 1979), and circadian rhythms of adrenal corticosterone content (Moore
and Eichler, 1974). Therefore, at least 2 distinct functional subpopulations of SCN axons,

controlling different endocrine outputs, exist.

91

92

In contrast to circadian regulation of endocrine function, which can be abolished by
selective interruption of axonal projections from the SCN, more severe disruption of
"suprachiasmatico-fugal" projections is required to abolish behavioral circadian rhythms.
Thus, complete isolation of an "island" of tissue containing the SCN (Stephan and Nunez,
1977); or knife cuts that partially isolate the SCN, severing axons that project dorsally,
laterally, and caudally, but not rostrally from the SCN (Nunez and Stephan, 1977) abolish
behavioral circadian rhythms. In contrast, more secective knife cuts placed in bilateral
parasagittal planes lateral to the SCN (Nunez and Stephan, 1977), a coronal plane either
rostral (Nunez and Stephan, 1977) or caudal to the SCN (Nishio, et al., 1979; Nunez and
Casati, 1979; Nunez and Stephan, 1977), or a horizontal plane dorsal to the SCN
(Experiment 1; Brown and Nunez, 1986; Eskes and Zucker, 1985) fail to affect behavioral
rhythms.

Regarding the functional role of the PVN and paraventriculo—spinal projections,
several interesting questions are raised by the present (and related) findings. It is very
difficult to resist the temptation to ask the teleological question: Why do 2 bundles of
paraventriculo—spinal projections exist? Related questions that are easier to answer
empirically include:

Are the fibers of bundles l and 2 collateral branches of the same parent axons 7

Do bundles l and 2 arise from distinct functional and/or anatomical subnuclei ?

What functions does bundle 2 serve ?

In addition to mediating photoperiodic regulation of the pineal gland and
reproduction, what other functions does bundle I serve ?

Is there a distinct subpopulation of PVN neurons that participates in photoperiodic
control of reproduction ?

While the answers to these questions are currently unknown, future research may
provide some of the answers. As mentioned above, evidence exists that bundle 1 might

mediate the feeding behavior that is elicited by microinjections of norepinephrine into the

93

PVN (Weiss and Leibowitz, 1986) and bundle 2 may be involved in the regulation of
energy balance (Gold, et al., 1977). Another question that merits investigation is whether
the PVN and bundles 1 and 2 are involved in nonreproductive responses to photoperiod.
Exposure of Syrian hamsters, and other mammalian species, to short days results in a
constellation of adaptations in addition to changes in reproductive capacity. Many of these
adaptations enhance the ability of the animal to survive the "real world" conditions of low
ambient temperatures and dwindling food supplies during winter. Examples of seasonal
cycles in non-reproductive functions which are dependent on photoperiod and the pineal
gland include: the shedding of antlers by deer during winter and spring months (Gross,
1969; Brown, et al., 1978); thyroid activity (Vriend, and Reiter, 1977; Vriend, et al.,
1979, 1982); fur color and fur density in Siberian hamsters (Duncan and Goldmanl984a,
1984b; Duncan, et al., 1985; Hoffman, 1978).

Photoperiodic responses that are not dependent on the pineal also exist. Hamsters
show adaptations that are dependent on photoperiod, directly related to energy balance,
and independent of the pineal. Exposure of Siberian hamsters to cold ambient temperatures
results in increased thermogenesis (as measured by resting 02 consumption). Siberian
hamsters kept in short days show enhanced capacity for thermogenesis, and animals that
have been kept in short days for 2 months are able to maintain normal body temperature
during subsequent exposure to ambient temperatures of -41° C (Heldmaier, et al., 1981).
Syrian hamsters housed in short days exhibit increases in body weight gain, carcass lipid
content, feed efficiency (i.e., weight gain per Kcal of food ingested), thermogenic capacity
(i.e., norepiniphrine-stimulated increase in 02 consumption), and show growth of brown
fat pads (Bartness and Wade, 1984), a major site thermogenesis (Foster and Frydman,
1978). These effects are exaggerated in hamsters fed a high fat diet (Bartness and Wade,
1984). These effects of short days are mimicked by peripheral injections of melatonin
given to hamsters housed in long days but, suprisingly, are not prevented by pinealectomy

(Bartness and Wade, 1984). Increased feed efficiency and body weight gain are also seen

94

in hamsters fed a high fat diet (Wade, 1982) and housed in long days, and the effects of
photoperiod and diet on these measures are greater in female than in male hamsters (Wade,
1983). Recent evidence suggests that the PVN may be involved in the effects of diet and
photoperiod on energy balance and thermogenesis since electrolytic lesions of the PVN
enhance high fat diet-induced, but prevent short photoperiod-induced obesity in hamsters
(Bartness, et al., 1985).

Thus, some commonalities (e.g., projections fiom the retina to the SCN and from the
SCN to the PVN) may exist in the neural pathways that regulate reproductive and those
that mediate nonreproductive responses to photoperiod However, there must also exist
differences between these pathways since metabolic responses to photoperiod are
independent of the pineal gland. In conclusion, while the neural pathways that mediate
pineal-dependent photoperiodic regulation of reproduction have been a "hotbed" of
scientific investigation in recent years, the neural pathways that mediate photoperiodic
regulation of thermogenesis and body energy balance and other nonreproductive
responses, some of which may be pineal independent ('2) remain largely uncharted

territory.

LIST OF REFERENCES

LIST OF REFERENCES

Ariens Kappers, J. The development of topographical relations and innervations of the
epiphysis cerebri in albino rat. ZZellforseh 52: 163-215, 1960.

Armstrong, W. E., S. Warach, G I Hatton, and T. H. McNeil]. Subnuclei 1n the rat
hypothalamic paraventricular nucleus: a cytoarchitectural, horseradish peroxidase and
immunocytochemical analysis. W 5: 1931-1958,1980.

Avrith, D. and G. Morgenson. Reversible hyperphagia and obesity following intracerebral
microinjections of colchicine into the ventromedial hypothalamus of the rat. W
155: 99-107, 197 8.

Bartness, T.J., E.L. Bittman, and G.N. Wade. Paraventricular nucleus lesions exaggerate
dietary obesity but block Photoperiod-induced weight gains and suspension of estrous
cyclicity in syrian hamsters. W 14: 427-430, 1985.

Bartness, T. J. and G. N. Wade. Photoperiodic control of body weight and energy
metabolism in Syrian hamsters (Mesocricetus auratus): role of pineal gland, melatonin,

gonads, and diet.End_chinQLngl_14:492-498, 1984.

Berk, ML. and J .A. Finkelstein. An autoradiographic determination of the efferent
projections of the suprachiasmatic nucleus of the hypothalamus. 013111395 226: 1-13,
198 1.

Boneau, C.A. The effects of violations of assumptions underlying the 1 test. EsychoLBull
51: 49-64, 1960.

Brown, M.H. Hypothalamic regulation of thermogenesis: Stimulation of the
paraventricular nucleus elicit thermal responses in brown adipose tissue. 5911921119591
mm: 1453, 1986.

Brown, M.H. and AA. Nunez. Hypothalamic circuits and circadian rhythms: effects of
knife cuts vary with their placement within the suprachiasmatic area. W11,
15: 705-711, 1986.

Brown, R. D. R. L. Cowan and J. F. Kavanaugh. Effect of pinealectomy on seasonal
androgen titers, antler growth and feed intake in white-tailed deer. LAnimaLSgi 41:
435-440, 1978.

Buijs, R. M. and P. Pevet. Vasopressin— and oxytocin-containing fibers in the pineal gland
and subcommissural organ of the rat. W 205: 11-17, 1980.

Campbell, C. S., J. Tabor, and J. Davis. Small effect of brown adipose tissue and major

effect of photoperiod on body weight in hamsters (Mesocricetus amatus). Bhysjol
32113130: 349- 352,1983.

95

96

Conrad, L.C.A. and D.W. Pfaff. Efferents from medial basal forebrain and hypothalamus
in the rat 11: an autoradiographic study of the anterior hypothalamus. W
162: 221-262, 1976.

Dahlstrom, A. Effect of colchicine on transport of amine storage granules in sympathetic
nerves of rat. W5: 111-113, 1968.

Dafny, N. Evidence that the rat pineal has direct neuronal connections via the pineal stalk.
W22: 858-861, 1983.

Dohanich, GP, and LG. Clemens. Brain areas implicated in cholinergic regulation of
sexual behavior. W15: 157-167, 1981.

Don Carlos, L. L. and J. A. Finkelstein. Hypothalamic projections to the spinal cord 1n the
hamster. W8: 218,1982.

Duncan, MJ. and B.D. Goldman. Hormonal regulation of the seasonal pelage color cycle
in the Djungarian hamster, W. 1. Role of the gonads and pituitary. 1
W 2le 89-95, 1984a.

Duncan, M.J. and B.D. Goldman. Hormonal regulation of the seasonal pelage color have
different critical daylengths in the Djungarian hamster, W. 1. Role of
prolactin. W 230: 97-103, 1984b.

Duncan, M. J, B. D. Goldman, M. N. DiPinto and M. H. Stetson. Testis function and
pelage color cycle 1n the Djungarian hamster, W. 1. Role of
prolactin. Endocrinology 116:424-430,1985.

Elliott, J. A. Circadian rhythms and photoperiodic time measurement in mammals. Em
W121 215: 2339-2346. 1976.

Ellis, G. B. and F. W. Turek. Time course of the photoperiod-induced change 1n sensitivity
of the hypothalamic-p ituitary axis to testosterone feedback 1n castrated male hamsters.

W194: 625-630 1979

Eskes G. A. and B. Rusak. Horizontal knife cuts in the suprachiasmatic area prevent
gonadal responses to photoperiod. W 61: 261-266, 1985.

Eskes, G. A. and I Zucker. Photoperiodic regulation of the hamster testis: Dependence on
circadian rhythms. W15: 1034-1038, 1978.

Foster, D. O. and M. L. Frydman. Nonshivering thermogenesis 1n the rat. 11.
Measurements of blood flow with microspheres point to brown adipose tissue as the
dominant site of the calorigenesis induced by noradrenaline. MW
56: 110-122, 1978

 

Freeman, M. E., M. S. Smith, S. J. Nazian and J. D. Neill. Ovarian and hypothalamic
control of the daily surges of prolactin during pseudopregnancy 1n the rat.

Endocrinnlmziz875-882, 1974.

Gold, R. M, A. P. Jones, P. E. Sawchenko and G. Kapatos. Paraventricular area: Critical
focus for a longitudinal neurocircuitry mediating food intake. Way 1821111-
1 119, 1977.

97

Gross, R. Photoperiodic control of antler cycles 1n deer. I. Phase shift and frequency
changes. W 110: 311 -,324 1969.

Guerillot, C., A. Pfister, J. Muller, and C. Da Lage. Recherche de l'origine des fibers
nerveuses extraorthosympathiques innervant l'epiphyse du rat (etude du tran

retrograde de la peroxydase de raifort). W 22: 371-378, 1982.

Hastings, M. H, A. C. Roberts, and J. Herbert. Neurotoxic lesions of the anterior
hypothalamus disrupt the photoperiodic but not the circadian system of the Syrian
hamster. W 40:316-324, 1985a.

Hastings, M.H., P. Winn, and SB. Dunnett. Neurotoxic amino acid lesions of the lateral
hypothalamus: A parametric comparison of the effects of ibotenate, N-methyl-D,L-
aspartate and quisqualate in the rat. M11325 360: 248-256, 1985b.

Hastings, M. H. and J. Herbert. Neurotoxic lesions of the paraventriculo-spinal projection
block the nocturnal rise in pineal melatonin synthesis in the Syrian hamster. Ngmacj
1&1! 62: 1-6, 1986.

Hatton, G.I., P. Cobbett, and AK. Salm. Extranuclear axon collaterals of paraventricular
neurons in the rat hypothalamus: Intracellular staining, immunocytochemistry and
electrophysiology. W 14: 123-132, 1985.

Hoffman, K. Effects of short photoperiod on puberty, growth, and molt in the Djungarian
hamster (Ehndomrnunms). Lflenrndflrt 5.4.: 29-35. 1978.

Hoffman, K. The influence of photoperiod and melatonin on testis size, body weight, and

pelage colour 1n the Djungarian hamster Wm),
267-282,1973.

Inouye, S.-I.T. and F.W. Turek. Horizontal knife cuts either ventral or dorsal to the
hypothalamic paraventricular nucleus block testicular regression in golden hamsters
maintained in short days. 813111363, 810: 102-107, 1986.

Kirk, R.E. Experimental Design: Procedures for the Behavioral Sciences. Belmont, CA:
Brooks/Cole, 1968.

Klein, M.N., R. Smoot, J.L. Weller, S. Higa, S.P. Markey, G.J. Creed, and D.M.
J acobowitz. Lesions of the paraventricular nucleus area disrupt the suprachiasmatic-
spinal cord circuit in the melatonin rhythm generating system. W 10: 647-
652, 198 3.

Korf, H.W. and U. Wagner. Evidence for a nervous connection between the brain and the
pineal organ in the guinea pig. W 202: 505-510, 1980.

Lehman, M.N., E.L. Bittman, and S. Winans-Newman. Role of the hypothalamic
paraventricular nucleus in neuroendocrine responses to daylength in the golden

hamster. 313111363 808: 25-32, 1984.

Leibowitz, S. F. Paraventricular nucleus: A primary site mediating adrenergic stimulation
of feeding anddrinking. WW8: 163-175, 1978.

98

Luiten, P.G.M., G.J. ter Horst, H. Karst, and AB. Steffens. The course of
paraventricular hypothalamic efferents to autonomic structures in medulla and spinal

cord. 312111323522: 374-378, 1985.

Malsbury, C. W. A schematic stereotaxic atlas of the golden hamster brain. QM
Wm 1:67,1977.

Matt, KS. and M.H. Stetson. Hypothalamic-pituitary-gonadal interactions during
spontaneous testicular recrudescence in golden hamsters. W 20: 739-746,
1979.

Mesulam, M.-M., E. Hegarty, H. Barbas, K.A. Carlson, E.C. Gower, A.G. Knapp,
M. B. Moss, and E. J. Mufson. Additional factors influencing sensitivity in the
tetramethyl benzidine method for horseradish peroxidase netn'ohistochemistry. .1

W28: 1255- 1259 1980

Moller, M. and H. W. Korf. The origin of central pinealopetal nerve fibers 1n the
Mongolian gerbil as demonstrated by the retrograde transport of horseradish
peroxidase. W 250:273-287,1983.

Moore, R.Y. and V. B. Eichler. Loss of a circadian adrenal corticosterone rhythm
following suprachiasmatic lesions in the rat. W63 42: 201-206, 1974.

Moore, R.Y. and DC. Klein. Visual pathways and the central neural control of a circadian
rhythm in pineal serotonin N-acetyltransferase activity. W 11: 17-33, 1974.

Moore. R.Y. and NJ. Lenn. A retinohypothalamic projection in the rat. W
146: 1-14, 1972.

Nance, D. W. Frontolateral hypothalamic knife cuts: Interruption of paraventricular
efferents to the thoracolumbar cord. W 1: 581-584, 1981.

Nishio, T.S., S. Shiosaka, H., H. Nakagawa, T. Sakumoto and K. Satoh. Circadian
feeding rhythm after hypothalamic knife-cut isolating suprachiasmatic nucleus. 21135191
33113123: 763-769, 1979.

Nunez, A.A. and M.J. Casati. The role of efferent connections of the suprachiasmatic in
the control of circadian rhythms. W 25: 263-267, 1979

Nunez, A.A. and PK. Stephan. The effects of hypothalamic knife cuts on drinking
rhythms and the estrous cycle of the rat. 3611318161 20: 224-234, 1977.

Olney, J. W. and M. T. Price. Excitotoxic amino acids as neuroendocrine research tools.

Mmhadsjnfinzymnlmlflit 379- 392 1983

Peterson, GM. and R.Y. Moore. Selective effects of kainic acid on diencephalic neurons.

mm: 202: 165-182, 1980.

Pickard GE. and A.J. Silvennan. Direct retinal projections to the hypothalamus, pirifonn
cortex and accessory optic nuclei in the golden hamster as demonstrated by a sensitive

anterograde horseradish technique. W 126:155-172, 1981.

99

Pickard, G. E. and F. W. Turek. The hypothalamic paraventricular nucleus mediates the
photoperiodic control of reproduction but not the effects of light on the circadian
rhythm of activity. W 48: 67- 72, 1983.

Rando, T.A., C.W. Bowers, and RE. Zigmond. Localization of neurons in the rat spinal
cord which project to the superior cervical ganglion. W 126:73—83, 1981.

Reiter, R J. and R. J. Hester. Inten'elationships between the pineal gland, the
superiorcervical ganglion, and the photoperiod 1n the regulation of the endocrine
systems of hamsters. W12: 1168- 1170, 1966.

Reiter, RJ., B.A. Richardson, and TS. King. The pineal gland and its indole products:
Their importance in the control of reproduction in mammals. In: The Pineal Gland,
edited by R. Relkin. New York: Elsevier Biomedical, 1983.

Rusak, B. and LP Morin. Testicular responses to photoperiod are blocked by lesions of
the suprachiasmatic nucleus in the golden hamster. W 15: 366-374, 197 6.

Rusak, B. and I. Zucker. Neural regulation of circadian rhythms. PhysigLRex 52: 449-
526, 1979.

Scouten, C.W., C.F. Cegavske, and L. Rozboril. A versatile carrier for simply
constructed wire knives and microsyringes. Way 26: 1115-1119, 1981.

Silverman, A. J, D. L. Hoffman, and E. A. Zimmerman. The descending afferent
connections of the paravenuicular nucleus of the hypothalamus (PVN).B;ain_Rgs_Bu11
6. 47- 61, 1981.

Steger, R. W., K. Matt, and A. Bartke. Neuroendocrine regulation of seasonal

reproductive activity in the male golden hamster. W 2: 191-201,
1985.

Stephan F.K., K.J. Berkley, and R.L. Moss. Efferent connections of the rat
suprachiasmatic nucleus. Wimp; 6: 2625-2641, 1981.

Stephan, F. K. and A. A. Nunez. Elimination of circadian rhythms in drinking, activity,
sleep, and brain temperature by isolation of the suprachiasmatic nuclei. Bfihflm 20:
1- 18,1977.

Stephan, F. K. and I. Zucker. Circadian rhythms 1n drinking behavior and locomotor
activity of rats are eliminated by hyothalamic lesions. W 62:
1583- 1586, 1972.

Stetson, M.H., M. Watson-Whitmyre, and KS. Matt. Termination of photorefractoriness
in golden hamsters-photoperiodic requirements. W 212: 81-88, 1977.

Swanson, L.W. and H.G.J.M. Kuypers. A direct projection from the ventromedial
nucleus and retrochiasmatic area to the medulla and spinal cord of the rat. m
L613 11: 307-312, 1980a.

Swanson, L.W. and H.G.J.M. Kuypers. The paraventricular nucleus of the
hypothalamus: cytoarchitectonic subdivisions and the organization of projections to the
pituitary, dorsal vagal complex and spinal cord as demonstrated by retrograde
fluorescence double labelling methods. W 124: 555-570, 1980b.

100

Tamarkin, L., J .S. Hutchison, and B.D. Goldman. Regulation of serum gonadotropins

by photoperiod and testicular hormone in the Syrian hamster. W 22:1528-
1533, 197 6.

Tulp, O.L., R. Frink, and E. Danforth. Effects of cafeteria feeding on brown and white
adipose tissue cellularity, thermogenesis, and body composition in rats. my]: 112:
2250-2260, 1982.

Turek, F. W. and G. B. Ellis. Steroid-dependent and steroid-independent aspects of the
photoperiodic control of seasonal reproductive cycles 1n male hamsters. In:
WW3. edited by B. K. Follett and D. E Follet.
Bristol, England: John Wright and Sons, 1981, pp. 251- 260.

Vriend, J. and R.J. Reiter. Free thyroxiod index in normal, melatonin treated and blind
hamsters. WW 2: 231-234., 1977.

Vriend, J., R.J. Reiter and GR. Anderson. Effects of the pineal and melatonin on thyroid
activity of male golden hamsters. W88: 189-195., 1979.

Vriend, J., B. A Richardson, M. K. Vaughn, L. Y. Johnson and R..J Reiter. Effects

melatonin on thyroid physiology of female hamsters. W 85: 79- 85.,
1982.

Wade, G.N. Obesity without overeating in golden hamsters. W 22: 701-707,
1982.

Wade, G. N. Dietary obesity 1n golden hamsters: reversibility and effects of sex and
photoperiod. W312: 131- 137, 1983.

Weiss, GR, and SP. Leibowitz. Efferent projections from the paraventricular nucleus
mediating az-noradrenergic feeding. Brain R25 31: 225-238, 1985.

Willis, G.L. and G.C. Smith. Short-term effects of intrahypothalamic colchicine. Bran
365216: 119-125, 1983.

Youngstrom, T. G. and A. A. Nunez. Anatomical features of the hypothalamic circuits that

mediate responses to photoperiod on the golden hamster. W 12: 212,
1986.

Zhang, T.-X. and J. Ciriello. Kainic acid lesions of paraventricular nucleus neurons
reverse the elevated arterial pressure after aortic baroreceptor denervation in the rat.

W858: 334-338, 1985.