11
Cell Tissue Res (1985) 242:301-311 cen and T't ae Research Springer-Verlag1985 Ependymoneuronal specializations between LHRH fibers and cells of the cerebroventricular system Gerald P. Kozlowski and Penelope W. Coates Department of Physiology, University of Texas Health Science Center at Dallas, Southwestern Medical School, Dallas, Texas, USA; Department of Anatomy, Texas Tech University Health Sciences Center, School of Medicine, Lubbock, Texas, USA Summary. Light- and electron-microscopic immunocyto- chemistry (LM-ICC and EM-ICC) were used to visualize luteinizing hormone-releasing hormone (LHRH) in fibres associated with ventricular ependyma and tanycytes of the median eminence. LM-ICC suggests that LHRH fibers ap- pear to enter the third ventricle. However, with EM-ICC, LHRH fibers are in fact found within ependymal canaliculi formed by adjacent ependymal cells. The canaliculi contain other myelinated and unmyelinated axons in addition to immunoreactive LHRH fibers. Thin slips of ependymal and tanycyte processes project into the canaliculi and enclose axons to varying degrees. At the median eminence many LHRH fibers bend sharply downwards from their ventricu- lar course and travel with tanycytic processes towards their common destination - the perivascular space of the hypo- physial-portal vascular system. Here, EM-ICC reveals that LHRH fibers closely contact basal processes of tanycytes. Lateral processes from tanycytes form glioplasmic sheaths which surround some individual LHRH fibers. A few LHRH terminals contact the perivascular space directly but more often are separated from the perivascular space by intervening glia. It is hypothesized that: (1)glia of this re- gion responds to the physiological state of the animal and may determine the degree of LHRH secretion by varying the extent of glial investment of LHRH terminals; and (2) may play a role during development by providing direction and support for LHRH fibers similar to that described for radial and other glial cells. Key words: Luteinizing hormone-releasing hormone (LHRH) - Tanycytes - Glial cells - Ependyma - Median eminence- Rats, mice One of the most striking features of the luteinizing hor- mone-releasing hormone (LHRH)-containing fiber system is the close association of these fibers with ependymal cells of the cerebroventricular system. This association was first observed with light-microscopic immunocytochemistry (LM-ICC) by Baker etal. (1975). However, the true interependymal nature of ventricularly associated LHRH Send offprint requests to: Dr. Gerald P. Kozlowski, Department of Physiology, University of Texas Health Science Center at Dallas, Southwestern Medical School, 5323 Harry Hines Blvd., Dallas, TX 75235, USA fibers could only be demonstrated with electron-microscop- ic immunocytochemistry (EM-ICC) (Kozlowski and Hos- tetter 1978). Since the basis for this interpretation was lim- ited, we now present further observations from this labo- ratory's ongoing investigations of LHRH fibers, ependymal cells and tanycytes of the ventricular system. Preliminary results of this investigation have been previously reported (Coates and Kozlowski 1982). Materials and methods The procedures utilized for these studies were similar to those described in previous reports (Kozlowski et al. 1980; Kozlowski and Nilaver 1983). Long-Evans rats and Swiss- Webster male mice were administered 0.5 I.U. of sodium pentobarbital subcutaneously. Animals were perfusion- fxed via an intracardiac cannula at room temperature using a mixture of 3% paraformaldehyde and 2% glutaraldehyde in 0.1% phosphate buffer at pH 7.4. After removal of the brain, areas of interest were blocked out and fixed by immersion overnight. Blocks were washed in 0.1 M phos- phate buffer at pH 7.2. Fifty-gm serial sections were cut with a vibrating microtome. The sections were transferred into marked wells of microtiter plates and placed on an orbit shaker during incubations and washes. Sections were washed in Tris-buffer and treated with 0.5% H20 z in Tris- saline (0.1 M Tris buffer, 0.9% NaCI and 0.5% gelatin) for 30 rain to eliminate endogenous peroxidase. They were then washed in Tris-saline for 15 min. The Tris-saline was replaced with Tris A, i.e., Tris-saline and 0.5% Triton X- 100 at pH 7.2 for 15 min. Sections were then transferred to Tris B, i.e., Tris A plus 1% Milipore-filtered normal goat serum. Two anti-LHRH sera were utilized: Kerdelhu~ No. 4152-8 and No. 4150-11 at 1:1000 (Kerdelhu6 1978). The anti-LHRH antisera incubations were carried out for 16-24 h in 0.1 M Tris B buffer at 4 C. Use of liquid-phase immunoabsorption controls in which 1 gg of LHRH (Ab- bott Laboratories, Chicago, IL) was added to 1 ml of anti- serum in final dilution abolished staining. The primary anti- serum was then replaced by wash solutions with Tris A followed by Tris B for 15 rain each. Goat anti-rabbit IgG in Tris B was then substituted for the above wash solution and used at 1 : 100 for 60 rain. Sections were again washed in Tris A and Tris B prior to using PAP (peroxidase-an- tiperoxidase complex) at 1 : 100 in Tris B for 60 min. Fol- lowing 3 washes in Tris-saline, the sections were incubated

Ependymoneuronal specializations between LHRH fibers and cells of the cerebroventricular system

  • Upload
    igc

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Cell Tissue Res (1985) 242:301-311 c e n and T't ae R e s e a r c h �9 Springer-Verlag 1985

Ependymoneuronal specializations between LHRH fibers and cells of the cerebroventricular system Gerald P. Kozlowski and Penelope W. Coates Department of Physiology, University of Texas Health Science Center at Dallas, Southwestern Medical School, Dallas, Texas, USA; Department of Anatomy, Texas Tech University Health Sciences Center, School of Medicine, Lubbock, Texas, USA

Summary. Light- and electron-microscopic immunocyto- chemistry (LM-ICC and EM-ICC) were used to visualize luteinizing hormone-releasing hormone (LHRH) in fibres associated with ventricular ependyma and tanycytes of the median eminence. LM-ICC suggests that LHRH fibers ap- pear to enter the third ventricle. However, with EM-ICC, LHRH fibers are in fact found within ependymal canaliculi formed by adjacent ependymal cells. The canaliculi contain other myelinated and unmyelinated axons in addition to immunoreactive LHRH fibers. Thin slips of ependymal and tanycyte processes project into the canaliculi and enclose axons to varying degrees. At the median eminence many LHRH fibers bend sharply downwards from their ventricu- lar course and travel with tanycytic processes towards their common destination - the perivascular space of the hypo- physial-portal vascular system. Here, EM-ICC reveals that LHRH fibers closely contact basal processes of tanycytes. Lateral processes from tanycytes form glioplasmic sheaths which surround some individual LHRH fibers. A few LHRH terminals contact the perivascular space directly but more often are separated from the perivascular space by intervening glia. It is hypothesized that: (1)glia of this re- gion responds to the physiological state of the animal and may determine the degree of LHRH secretion by varying the extent of glial investment of LHRH terminals; and (2) may play a role during development by providing direction and support for LHRH fibers similar to that described for radial and other glial cells.

Key words: Luteinizing hormone-releasing hormone (LHRH) - Tanycytes - Glial cells - Ependyma - Median eminence- Rats, mice

One of the most striking features of the luteinizing hor- mone-releasing hormone (LHRH)-containing fiber system is the close association of these fibers with ependymal cells of the cerebroventricular system. This association was first observed with light-microscopic immunocytochemistry (LM-ICC) by Baker etal. (1975). However, the t rue interependymal nature of ventricularly associated LHRH

Send offprint requests to: Dr. Gerald P. Kozlowski, Department of Physiology, University of Texas Health Science Center at Dallas, Southwestern Medical School, 5323 Harry Hines Blvd., Dallas, TX 75235, USA

fibers could only be demonstrated with electron-microscop- ic immunocytochemistry (EM-ICC) (Kozlowski and Hos- tetter 1978). Since the basis for this interpretation was lim- ited, we now present further observations from this labo- ratory's ongoing investigations of LHRH fibers, ependymal cells and tanycytes of the ventricular system. Preliminary results of this investigation have been previously reported (Coates and Kozlowski 1982).

Materials and methods

The procedures utilized for these studies were similar to those described in previous reports (Kozlowski et al. 1980; Kozlowski and Nilaver 1983). Long-Evans rats and Swiss- Webster male mice were administered 0.5 I.U. of sodium pentobarbital subcutaneously. Animals were perfusion- fxed via an intracardiac cannula at room temperature using a mixture of 3% paraformaldehyde and 2% glutaraldehyde in 0.1% phosphate buffer at pH 7.4. After removal of the brain, areas of interest were blocked out and fixed by immersion overnight. Blocks were washed in 0.1 M phos- phate buffer at pH 7.2. Fifty-gm serial sections were cut with a vibrating microtome. The sections were transferred into marked wells of microtiter plates and placed on an orbit shaker during incubations and washes. Sections were washed in Tris-buffer and treated with 0.5% H20 z in Tris- saline (0.1 M Tris buffer, 0.9% NaCI and 0.5% gelatin) for 30 rain to eliminate endogenous peroxidase. They were then washed in Tris-saline for 15 min. The Tris-saline was replaced with Tris A, i.e., Tris-saline and 0.5% Triton X- 100 at pH 7.2 for 15 min. Sections were then transferred to Tris B, i.e., Tris A plus 1% Milipore-filtered normal goat serum. Two anti-LHRH sera were utilized: Kerdelhu~ No. 4152-8 and No. 4150-11 at 1:1000 (Kerdelhu6 1978). The anti-LHRH antisera incubations were carried out for 16-24 h in 0.1 M Tris B buffer at 4 C. Use of liquid-phase immunoabsorption controls in which 1 gg of LHRH (Ab- bott Laboratories, Chicago, IL) was added to 1 ml of anti- serum in final dilution abolished staining. The primary anti- serum was then replaced by wash solutions with Tris A followed by Tris B for 15 rain each. Goat anti-rabbit IgG in Tris B was then substituted for the above wash solution and used at 1 : 100 for 60 rain. Sections were again washed in Tris A and Tris B prior to using PAP (peroxidase-an- tiperoxidase complex) at 1 : 100 in Tris B for 60 min. Fol- lowing 3 washes in Tris-saline, the sections were incubated

Fig.

1.

Sec

tion

thr

ough

the

ros

tral

hyp

otha

lam

us s

how

ing

the

posi

tion

of

LH

RH

fi

bers

in

the

supr

achi

asm

atic

tra

ct (

SCT)

. T

here

are

num

erou

s L

HR

H

fibe

rs l

ater

ally

(ar

row

head

s).

III

thir

d ve

ntri

cle;

OC

opt

ic c

hias

m,

x 45

0

Fig.

2.

Sec

tion

thr

ough

the

re

gion

of

the

med

ian

emin

ence

of

the

hypo

thal

amus

. L

HR

H

fibe

rs a

re

clos

ely

asso

ciat

ed

wit

h ep

endy

mal

ce

lls

of t

he

thir

d ve

ntri

cle

(Ill

) an

d sw

eep

vent

rola

tera

lly

(arr

owhe

ads)

thr

ough

the

arc

uate

nuc

leus

tow

ards

hyp

ophy

sial

-por

tal

bloo

d ve

ssel

s,

x 45

0

Fig.

3a,

b.

S

ecti

on t

hrou

gh

the

caud

al

hypo

thal

amus

, a

Num

erou

s L

HR

H

fibe

rs a

re

asso

ciat

ed w

ith

epen

dym

al c

ells

of

the

thir

d ve

ntri

cle

(111

) as

wel

l as

th

e ve

ntra

l su

rfac

e of

th

e br

ain

(arr

ows)

. x

450.

b H

igh

mag

nifi

cati

on o

f th

e nu

mer

ous,

bea

ded,

LH

RH

fi

bers

in

the

epen

dym

al c

ell

laye

r,

x 78

0

303

in a freshly prepared mixture of 0.015% 3,3'-diaminobenzi- dine tetrahydrochloride and 0.003% H20 z in 0.1 M Tris buffer at pH 7.6. Sections were incubated until visible reac- tion product was noted in the median eminence. This step usually required 2-4 min. Sections were washed in Tris- saline 3 times at 10 rain each. Those intended for ultrastruc- tural examination were then washed in 0.1 M phosphate buffer, returned to 3% glutaraldehyde in phosphate buffer for 1 h. They were then examined with the light microscope. Selected areas of interest within the section were removed with a razor blade. These smaller (approximately 2-10 mm 2) pieces were washed again in phosphate buffer and postfixed in 1% OsO4 in 0.1 M phosphate buffer for 30min. They were then dehydrated, infiltrated, and embedded in Spurr's low-viscosity resin. Alternatively, other sections were embedded in Spurr's resin on a glass slide and examined for areas of interest. Small pieces were then removed, glued to a blank block and thin-sectioned. The ultrathin (silver to light gold) sections were stained with aqueous uranyl acetate and Reynold's lead citrate. Electron micrographs were taken on a Hitachi H-600 trans- mission electron microscope or a Zeiss EM 10.

Results

LM-ICC revealed extensive associations of L H R H fibers with the ependyma throughout the rostral-caudal extent of the third ventricle. Rostrally, there was a discrete tract of L H R H fibers overlying the optic chiasm termed the su- prachiasmatic tract (Fig. 1). At the level of the median emi- nence (MR), many L H R H fibers were associated with the ependymal lining of the ventricle, especially ventrally (Fig. 2). Here, L H R H fibers entered the adjacent subependymal neuropil and coursed ventrolaterally. Cau- dally, L H R H fibers maintained their close relationship (Fig. 3 a, b) with the ependymal layer. At the most caudal regions of the hypothalamus, L H R H fibers coursed in pa- rallel with ependyma (Fig. 4a, b). Fewer L H R H fibers were associated with ependyma of the cerebral aqueduct (Fig. 5) in comparison with the third ventricle.

EM-ICC for L H R H revealed that instead of entering the ventricular lumen, L H R H fibers travelled within ependymal tunnels (termed canaliculi) which were formed by apposing ependymal cells (Fig. 6). Ependymal canaliculi are similar to bile canaliculi and are formed by grooves in the lateral surface of adjacent cells. This forms the inside wall of the canaliculus. Finger-like slips of tanycyte cyto- plasm projected into the canaliculus (Figs. 7, 8 a, b, c) and partitioned off bundles of fibers, not all of which are immu- noreactive, which course within the canaliculus. A few non- reactive intracanalicular fibers were lightly myelinated (Figs. 7, 8 a).

At the level of the median eminence, the L H R H fibers abruptly turned ventrally and coursed parallel to the shafts of tanycytes. Many L H R H fibers were in direct contact with the basal processes of tanycytes (Figs. 9-11). Varicosi- ties of L H R H fibers were filled with immunoreactive neurosecretory granules with a mean diameter of 104.6nm_+3.6 (S.E.M.) (Fig. 10). Near the perivascular space, tanycytic glioplasmic sheaths partially or wholly enc-

losed L H R H fibers at their pre-terminal or terminal por- tions (Figs. 12a, b, 13a, b, c). These ependymoneuronal specializations are summarized diagrammatically in Fig. 14).

Discussion

With LM-ICC, many L H R H fibers are consistently found closely associated with ependyma of the third and other portions of the cerebroventricular system (Figs. 1-5). This observation has been used by some investigators (Bennett- Clarke and Joseph 1982; Jennes and Stumpf 1980; Burcha- nowski et al. 1979, for review see Knigge et al. 1976) to support the hypothesis that L H R H is released into the cere- brospinal fluid (CSF) for transport to the ME. One tenet of the ventricular-route hypothesis (for review see Koz- lowski 1982) is that L H R H neurons secrete L H R H directly into the CSF via intraventricular nerve endings. Yet, despite the rich investment of L H R H fibers to ependyma, as visual- ized with LM-ICC, at the ultrastructural level, we and other investigators (Krisch 1980; Paul Goldsmith, personal com- munication) have been unable to find supra-ependymal fibers or intraventricular nerve endings containing immuno- reactive LHRH. We therefore conclude that, if present, they must be few in number. This finding is consistent with the failure of Cramer and Barraclough (1975) and Coppings et al. (1977) to detect radioimmunoassayable L H R H in CSF of rat and sheep. In the human, Miyake et al. (1980) were unable to find L H R H in the CSF of 29 out of 36 wom- en, the remaining 7 subjects had less than 0.95 pg/ml. Thus, the functional significance for the proximity of L H R H fibers to the ventricular lumen may not be that of secretion of L H R H directly into CSF. However, we do not eliminate the possibility that diffusion of L H R H into the CSF could occur. Uemura et al. (1981) measured small amounts of L H R H in CSF from anesthetized diestrous rats with levels of 1.4-1.9 pg/lal as opposed to those reported by Knigge and his associates, 308_+ 16.2 pg/gl (Joseph et al. 1975), 140.6 + 16.6 pg/gg (Morris et al. 1975) and 69.3 +_ 25.9 pg/gl (as estimated by Uenmra et al. 1981 from the data in Morris and Knigge 1976). Recently, L H R H was found in CSF of the monkey third ventricle (M. Ferin, personal communi- cation). Apparently L H R H appears there in a pulsatile manner having a time-course similar to the pulsatile release of L H R H in portal blood (Carmel et al. 1976).

The exact nature of the relationship between L H R H fibers and ependyma is best elucidated with EM-ICC which shows that L H R H fibers course between ependymal cells in canaliculi. Millhouse (1972) and Brawer (1972) described profiles of unmyelinated axons and dendrites between ependymal cells. In addition, we found myelinated fibers also course between ependymal cells and tanycytes. It would be of great interest to determine if fibers which coexist with L H R H fibers in the canaliculi are functionally related.

There is considerable interest in the relationship between neurosecretory cells and specialized ependymoglial cells of the third ventricle. Synaptoid contacts between neurons and both tanycytes and astrocytes in the median eminence have been described (Gfildner and Wolff 1973 ; Knowles 1967;

304

Fig. 4a, b. Section through the dorso-caudal third ventricle. a LHRH fibers are closely associated with the third ventricle (II1). Most are running vertically (arrows) and parallel with the third ventricle. x450.

b Higher magnification reveals a vertical LHRH fiber coursing through the ependymal cell layer (arrow). x 1780

Fig. 5. Section through the midbrain. A few LHRH fibers (arrows) are associated with the ventral ependyma of the cerebral aqueduct (CA). x 450

Kobayashi et al. 1970). In our material , it was difficult to demonst ra te synaptoid contacts because (1) we did not combine E M - I C C with special stains for synaptoid con- tacts; and (2) the final react ion product from the EM-ICC technique masked the presynapt ic density characteristic of synaptoid contacts. Nevertheless, in lightly stained immunoposi t ive L H R H fibers there were suggestions that such contacts existed between L H R H fibers and tanycytes.

Future experiments could provide convincing evidence for their presence by using colloidal gold as the EM-ICC marker instead of PAP in combinat ion with special stains for synaptoid elements such as ethanolic-phosphotungst ic- acid (E-PTA) or bismuth-iodide-uranylacetate (BIUL). Theoretically, such contacts could play a role in which L H R H fibers could influence tanycytes to change cell shape and volume, or even alter the extent of the glioplasmic

Fig.

6.

Tan

ycyt

es (

N n

ucle

us)

of t

he t

hird

ven

tric

le.

Epe

ndym

al c

anal

icul

i (E

C)

are

inte

rpos

ed b

etw

een

adja

cent

tan

ycyt

es.

The

se c

anal

icul

i fo

rm t

unne

ls c

onta

inin

g L

HR

H

(arr

ows)

and

oth

er i

mm

unon

egat

ive

fibe

rs.

TJ ti

ght

junc

tion

, x

8600

Fig

. 7.

Hig

h m

agni

fica

tion

of

epen

dym

al c

anal

icul

i (E

C).

Sev

eral

non

-im

mun

orea

ctiv

e, u

nmye

lina

ted

fibe

rs (

arro

whe

ads')

, an

occa

sion

al t

hinl

y m

yeli

nate

d no

n-im

mun

orea

ctiv

e fi

ber,

an

d an

L

HR

H

fibe

r (a

rrow

) ar

e pr

esen

t.

Not

e th

e cl

ose

prox

imit

y of

the

se f

iber

s to

the

lum

en

on

the

thir

d ve

ntri

cle

whi

ch c

ould

acc

ount

fo

r th

eir

appe

arin

g to

be

wit

hin

the

vent

ricl

e at

the

LM

lev

el.

The

re a

re n

umer

ous

mem

bran

ous

exte

nsio

ns o

f ta

nycy

tes

proj

ecti

ng i

nto

the

epen

dym

al c

anal

icul

i. T

ypic

al j

unct

iona

l co

mpl

exes

are

pre

sent

at

the

apic

al

surf

ace,

x

18 0

00

306

coverings of L H R H terminals in the ME. In this way L H R H neurons could regulate the activity of tanycytes by uncovering the glioplasmic sheath, thereby assisting the release of L H R H into the perivascular space. Lichtensteiger et al. (1978) provided evidence that dopaminergic and choli- nergic mechanisms can cause variations in the extension of tanycyte end-feet in the neurohemal contact zone of the ME. Utilizing morphometric analysis of non-neuronal ele- ments (tanycytes) from the lateral part of the ME, they found that 60-70% of the basement membrane was occu- pied by tanycytes in sham-operated or saline-injected cont- rols. After administration of nicotine, there was an increase in the presence of uncovered neuronal profiles. Schiebler et al. (1978) reported results contrary to that expected for the secretion of L H R H under conditions of high release of neurohormone. They found that after castration, glial surface area in the zona externa increased by 23%, and after adrenalectomy by 32%. On the other hand, Witt- kowski and Scheuer (1974) showed that after bilateral adre- nalectomy the neuro-vascular contacts increased from 20 to 40%. This effect was reversible since application of an extract of tissue from ME of the stalk caused a decrease in neurovascular contacts to 15%.

In this study we rarely found L H R H terminals in direct contact with the perivascular space of hypophysial-portal capillaries (Figs. 12a, 13a, b). Most L H R H terminals were either distant from their expected site of termination (Fig. 12b) or separated from the perivascular space by thin glioplasmic sheaths (Fig. 13c) formed by tanycytes. We suggest that the amount of enclosure of LHRH terminals by glioplasmic sheaths relates to the reproductive state of the animal. We predict that such a picture would change drastically during the afternoon of proestrous or after cast- ration when increased L H R H release occurs. Studies of the relationship between magnocellular neurosecretory neurons and their glia provide a basis for such speculation. Tweedle and Hatton (1980) and Tweedle (1983) showed that special- ized glial cells (pituicytes) that enclose neurosecretory end- ings in the neurohypophysis retracted from around neurosecretory endings during parturition (Tweedle and Hatton 1982) thereby allowing for greater release of hor- mone when required. Astrocytic glial processes also retracted from between magnocellular neurosecretory neurons in the supraoptic nucleus and nucleus circularis from pregnancy to lactation (Hatton and Tweedle 1982; Slam et al. 1983) allowing greater cell-cell contact that may serve a synchronizing function, particularly in oxytocin cells participating in the milk ejection reflex.

Tanycytes may share some modulatory functions attrib- uted to other glial cells. Under conditions of high neuronal activity, glial cells take up extracellular potassium ions (K +) that may provide a buffering mechanism for K + concentra- tion in the extracellular space. Furthermore, some glial cells can also take up and store gamma aminobutyric acid (GABA) which can be released with elevated intracellular concentrations of K + (Martin 1981). Thus, like other glial cells, tanycytes may be sensitive to changes in extracellular K +. They may respond to increasing neuronal activity of nearby neurosecretory nerve endings of the ME by depola- rizing (due to an influx of extracellular K+), perhaps caus- ing changes in tanycyte cell shape. The change in cell shape may relate directly to the degree of investment of the

tanycyte covering neurosecretory nerve endings at the perivascular space.

Another possible glial-neuronal relationship of interest arises from the work of Lasek and Tytell (1981) who showed that glial cells in the squid transfer as much as 40% of their newly synthesized proteins, traversin and actin, to the giant axon. Although no evidence has been brought to light for ependymoneuronal transport, the close morphological associations suggest that ependymoneuronal interactions could include transport of material from ependymal cell to LHRH fibers.

The special relationship between LHRH nerve fibers and unspecialized (mural) and specialized (tanycyte) ependymal cells is similar to neuronoglial interactions dur- ing brain development. Rakic (1981) described the interac- tion between radial glial cells and immature neurons. He showed that radial glial cells play a crucial role in the orien- tation, displacement and positioning of neurons within the developing cerebral cortex. Radial glial cells, which usually have their somata near the ventricular surface, have an elongated process that traverses the entire width of the cere- bral wall to terminate at the pial surface (Rakic 1981). Like- wise, tanycytes usually have their somata near the ventricu- lar surface, and an elongated process (cf. "stretch-cell", Horstmann 1954) which traverses the ME to terminate on the perivascular spaces and pial surface. Similarly, both cell types may be involved in the transport of substances (Oksche 1968; Chu-Wang et al. 1981; Ivy and Killackey 1978). Another similarity can now be drawn on the close proximity between specialized glial cells and neurons such as that which exists for radial glial cells and cortical neurons, and between tanycytes and LHRH fibers. Cortical neurons migrate along processes of radial glial cells as a general principle of development in many brain areas such as Bergmann glial cells of the cerebellum and Miiller cells of the retina (Rakic 1971, 1972, 1975, 1981). Neurons reach their proper destination and establish correct synaptic rela- tions via this mechanism. Similarly, tanycytes may provide a substrate for LHRH fibers, giving them direction and guidance as they travel to their characteristic sites of termi- nation on portal capillaries of the ME. Our speculation with regard to the guidance and directionality provided to L H R H fibers by ependymal tunnels and tanycytes during development is further supported by the work of Singer and colleagues (the 'blueprint ' hypothesis, 1979; Nord- lander and Singer 1982), Silver and Sidman (1980) and Krayanek and Goldberg (1981). Fig. 14 summarizes diag- rammatically these ependymoneuronal specializations.

Acknowledgments. We thank Mrs. Jane Moore-Shwartz, Mr. Ant- hony Frisbie and Mr. Stephen Sickerman for their assistance in this project, and Mrs. Carol McLain for typing the manuscript. Supported by grants from the NIH: AA-06014 and HD-15040 (GPK) and HD-12833 (PWC); and a grant from the Inst. Biomed. Res., TTUHSC (PWC).

Fig

. 8a

--c.

Hig

h m

agni

fica

tion

of

seve

ral

exam

ples

of

epen

dym

al c

anal

icul

i, a

A s

mal

l m

embr

anou

s pr

oces

s of

the

tan

ycyt

e (a

rrow

) par

tial

ly c

ompa

rtm

enta

lize

s th

e im

mun

opos

itiv

e L

HR

H

fibe

r fr

om a

djac

ent

imm

unon

egat

ive,

mos

tly

unm

yeli

nate

d fi

bers

. •

b A

n L

HR

H

fibe

r (a

rrow

) adj

acen

t to

an

im

mun

oneg

ativ

e,

unm

yeli

nate

d fi

ber,

x

1600

0.

cA

tany

cyte

pro

vide

s an

ext

rem

ely

thin

ton

gue

of m

embr

ane

form

ing

a ca

nali

culu

s co

ntai

ning

an

imm

unop

osit

ive

LH

RH

fib

er (

arro

w).

Fig

. 9.

Hor

izon

tal

sect

ion

thro

ugh

the

med

ian

emin

ence

sho

win

g nu

mer

ous

imm

unop

osit

ive

LH

RH

fib

ers

in c

ross

-sec

tion

. T

he c

ross

-sec

tion

al d

iam

eter

of

each

fib

er v

arie

s ac

cord

ing

to w

heth

er t

he s

ecti

on o

ccur

s at

a v

aric

osit

y (w

here

the

dia

met

er w

ould

be

grea

test

) or

the

int

erva

rico

se s

egm

ent.

L

arge

, va

rico

se s

egm

ents

are

fil

led

wit

h ne

uros

ecre

tory

gra

nule

s.

Man

y fi

bers

are

adj

acen

t to

the

sha

ft o

f ta

nycy

tes

(7-)

. x

7200

Fig

. 10

. L

on

git

ud

inal

sec

tion

th

rou

gh

th

e m

edia

n e

min

ence

sho

ws

a va

rico

se p

ort

ion

of

an L

HR

H f

iber

bet

wee

n t

wo

sha

fts

of

tan

ycy

tes

(7).

The

im

mu

no

po

siti

ve

LH

RH

neu

rose

cret

ory

gr

anul

es (

NSG

) hav

e a

mea

n d

iam

eter

of

104.

6 n

m •

3.6

(S.E

.M.)

. x

22 5

00

Fig

. 11

. a

Ho

rizo

nta

l se

ctio

n th

rou

gh

the

med

ian

em

inen

ce d

emo

nst

rate

s L

HR

H

fibe

rs i

n di

rect

ap

po

siti

on

to

the

bas

al p

roce

sses

of

tan

ycy

tes

(7).

x

1750

0.

b L

on

git

ud

inal

sec

tion

d

emo

nst

rate

s a

vari

cose

seg

men

t o

f an

LH

RH

fib

er i

n ap

po

siti

on

to

the

tan

ycy

te (

7)

proc

ess,

x

25 5

00.

c L

on

git

ud

inal

sec

tion

dem

on

stra

tes

rela

tio

nsh

ip o

f L

HR

H f

iber

s to

a t

any

cyte

an

d o

ther

unr

eact

ive

fibe

rs,

x 23

000

Fig

. 12

. a

Low

mag

nifi

cati

on o

f th

e zo

na e

xter

na o

f th

e m

edia

n em

inen

ce s

how

s ab

un

dan

t L

HR

H

fibe

rs t

erm

inat

ing

near

fen

estr

ated

cap

illa

ries

(C

) of

the

hyp

ophy

sial

por

tal

syst

em.

x 35

00.

b H

ighe

r m

agni

fica

tion

of

the

area

out

line

d by

the

mid

dle

box

of F

ig.

12a

reve

als

that

tan

ycyt

es (

7) f

orm

sl

ips

of c

ytop

lasm

whi

ch p

arti

ally

enc

lose

a

nu

mb

er o

f L

HR

H

fibe

rs.

• 15

500

Fig

. 13

. a

Hig

her

mag

nifi

cati

on o

f en

clos

ed a

rea

* of

Fig

. 12

a re

veal

s an

LH

RH

-fib

er

abut

ting

dir

ectl

y on

the

bas

al l

amin

a (B

L) w

itho

ut a

ny i

nter

veni

ng g

lial

cyt

opla

sm,

x 23

000.

b

Hig

her

mag

nifi

cati

on o

f en

clos

ed a

rea

** o

f F

ig.

12a

show

s an

othe

r ex

ampl

e of

an

LH

RH

-fib

er t

erm

inat

ing

dire

ctly

on

the

basa

l la

min

a. N

ote

othe

r L

HR

H-f

iber

s an

d no

n-re

acti

ve

fibe

rs a

ssoc

iate

d w

ith

thin

tan

ycyt

e pr

oces

ses,

x

2150

0. c

An

LH

RH

-fib

er h

avin

g an

int

erve

ning

sli

p of

gli

al c

ytop

lasm

bet

wee

n it

and

the

bas

al l

amin

a,

x 16

500

310

A

"~OVLI"

�84

..~

.......... 5

Fig. 14. Summary diagram illustrating the relationship between LHRH fibers and specialized ependyma of the ventricular system. 1 The course of an LHRH fiber related to the ventricular system. The cell body of origin is located in the medial preoptic area at the level of the organum vasculosum lamina terminalis (OVLT). Its process extends caudo-ventrally to the zona externa of the median eminence. MB mammillary body. 2 At the level of the optic chiasm (OC), as indicated by 1A, the LHRH fiber enters the ependymal lining of the third ventricle and travels between ependymal cells (dotted lines). 3 LHRH fibers (beaded fiber) as well as other fibers, travel within ependymal canaliculi. 4 At the level of the median eminence (ME) the fibers, (as indicated by 1B) leave the ependymal layer to travel with the shafts of tanycytes (T) toward capillaries of the hypophysial portal plexus. 5 LHRH fibers are closely apposed to tanycyte shafts and can be covered by slips of glioplasm (arrow). At the perivascular space (PVS) glioplasmic sheaths may cover, partially cover or expose LHRH terminals. This would serve to impede or enhance the secretion of LHRH when required

References

Baker BL, Dermody WC, Reel JR (1975) Distribution of gonado- tropin-releasing hormone in the rat brain as observed with im- munocytochemistry. Endocrinology 97:125-135

Bennett-Clarke C, Joseph SA (1982) Immunocytochemical distri- bution of LHRH neurons and processes in the rat: hypothalam- ic and extrahypothalamic locations. Cell Tissue Res 221 : 493-504

Brawer JR (1972) The fine structure of the ependymal tanycytes at the level of the arcuate nucleus. J Comp Neurol 145:25~42

Burchanowski BJ, Knigge KM, Sternberger LA (1979) Rich ependymal investment of lutcoliberin (LHRH) fibers revealed

immunocytochemically in Golgi-like image. Proc Natl Acad Sci USA 76:6671 6674

Carmel PW, Araki S, Forin M (1976) Pituitary stalk portal blood collection in Rhesus monkeys; evidence for pulsatile release of gonadotropin-releasing hormone (GnRH) Endocrinology 99:243 248

Chu-Wang I-W, Oppenheim RW, Farel P (1981) Ultrastructure of migrating spinal motoneurons in anuran larvae. Brain Res 213:307-318

Coates PW, Kozlowski GP (1982) Immunoreactive LHRH-fibers are intimately associated with tanycytes. Soc Neurosci Abs 8:324

Coppings RJ, Malvan PV, Ramirez VD (1977) Absence of immu-

311

noreactive luteinizing hormone-releasing hormone in ovine ce- rebrospinal fluid collected from the third ventricle. Proc Soc Exp Biol Med 154:219-223

Cramer OM, Barraclough CA (1975) Failure to detect luteinizing hormone-releasing hormone in third ventricle cerebral spinal fluid under a variety of experimental conditions. Endocrinology 96 : 913-921

Giildner FH, Wolff JR (1973) Neuro-glial synaptoid contacts in the median eminence of the rat : ultrastructure, staining proper- ties and distribution on tanycytes. Brain Res 61:217 234

Hatton GI, Tweedle CD (1982) Magnocellular neuropeptidergic neurons in hypothalamus: increases in membrane apposition and number of specialized synapses from pregnancy to lacta- tion. Brain Res Bull 8:197-204

Horstmann E (1954) Die Faserglia des Selachiergehirns. Z Zell- forsch 39: 588-617

Ivy GW, Killackey HP (1978) Transient populations of glial cells in developing rat telencephalon revealed by horseradish peroxi- dase. Brain Res 158 : 213-218

Jennes L, Stumpf WE (1980) LHRH-systems in the brain of the golden hamster. Cell Tissue Res 209:23%256

Joseph SA, Sorrentino SJ, Sundberg DK (1975) Releasing hor- mones, LRF and TRF, in the cerebrospinal fluid of the third ventricle. In: Brain-endocrine interaction II. The ventricular system. 2nd Int Symposium, Shiyuoka 1974, Karger, Basel, pp 306-312

Kerdelhu6 B (1978) Some immunological aspects of recently identi- fied neurosecretory peptides with special reference to thyro- tropin releasing hormone (TRH) and luteinizing hormone (LHRH). In: Vincent JD, Kordon C (ed) Biologic cellulaire des processus neurosecretories hypothalamiques. Colloques In- ternationaux du CNRS No 280, Paris, pp 289-307

Knigge KM, Joseph SA, Sladek JR, Notter MF, Morris M, Sundberg DK, Holzwarth MA, Horrman GE, O'Brien L (1976) Uptake and transport activity of the median eminence of the hypothalamus. Int Rev Cytol 45 : 383-408

Knowles F (1967) Neuronal properties of neurosecretory cells. In: Stutinsky F (ed) Neurosecretion. Springer, Berlin, pp 8-19

Kobayashi H, Matsui T, Ishii S (1970) Functional electron micros- copy of the hypothalamic median eminence. Int Rev Cytoi 29:281-381

Kozlowski GP (1982) Fine structural features of immunolabeled neurons containing luteinizing-hormone releasing hormone, va- sopressin, and neurophysin: In: Farner DS, Lederis K (eds) Neurosecretion: Molecules, cells, systems. Plenum Pub Co, New York, pp 61-70

Kozlowski GP, Hostetter G (1978) Cellular and subcellular local- ization and behavioral effects on gonadotropin-releasing hor- mone (Gn-Rh) in the rat. In: Scott DE, Kozlowski GP, Weindl A (eds) Brain-endocrine interaction. III. Neural hormones and reproduction, 3rd International Symposium, Wfirzburg, Germany. Springer-Karger, Basel, pp 138-153

Kozlowski GP, Nilaver G (1983) Immunoelectron microscopy of neuropeptides: theoretical and technical considerations. In: Barker JL, McKelvy JF (eds) Current methods in cellular neurobiology, Vol 1 : Anatomical techniques. John Wiley und Sons, NY, pp 133-174

Kozlowski GP, Brownfield MS, Hostetter G (1978) Neurophysin and gonadotropin-releasing hormone (Gn-RH) containing hypothalamo-fugal neurosecretory system. In : Lederis K, Veale WL (eds) Current studies of hypothalamic function, 1978. Part I: Hormones. Springer, Karger, Basel, pp 15-26

Kozlowski GP, Chu L, Hostetter G, Kerdelhue B (1980) Cellular characteristics of immunolabeled luteinizing hormone releasing hormone (LHRH) neurons. Peptides I:37-46

Krayanek S, Goldberg S (1981) Oriented extracellular channels and axonal guidance in the embryonic chick retina. Dev Biol 84:41-50

Krisch B (1980) Immunocytochemistry of neuroendocrine systems. Prog Histochem Cytochem 13:1 163

Lasek RJ, Tytell MA (1981) Macromolecular transfer from glia to the axon. J Exp Biol 95:153-165

Lichtensteiger W, Richards JG, Kopp HG (1978) Possible partici- pation of non-neuronal elements of median eminence in neuro- endocrine effects of dopaminergic and cholinergic systems. In : Brain endocrine interaction Ill. Neural hormones and repro- duction. 3rd Int Symp, Wiirzburg 1977, Karger, Basel, pp 251 262

Martin JH (1981) Properties of cortical neurons, the EEG, and the mechanisms of epilepsy. In : Kandel ER, Schwartz JH (eds) Principles of neural science. Elsevier, NY, p 465

Millhouse OE (1972) Light and electron microscopic studies of the ventricular wall. Z Zellforsch 127:149-174

Miyake A, Kurachi H, Kawamura Y, Aono T, Kurachi K, Yana- gida T (1980) Luteinizing hormone-releasing hormone in cere- brospinal fluid of women. Endocrinol Jpn 27:117-119

Morris M, Knigge KM (1976) The effects of pentobarbital and ether anesthesia on hypothalamic LH-RH in the rat. Neuroendocrinology 20:193-200

Morris M, Tandy B, Sundberg DK, Knigge KM (1975) Modifica- tion of brain and CSF LH-RH following deafferentation. Neuroendocrinology 18 : 131-135

Nordlander RH, Singer M (1982) Spaces precede axons in Xenopus embryonic spinal cord. Exp Neurol 75:221-228

Oksche A (1968) Die pr/inatale und vergleichende Entwick- lungsgeschichte der Neuroglia. Acta Neuropathologica Suppl 4:4-19

Rakic P (1971) Neuron-glia relationship during granule cell migra- tion in developing cerebellar cortex. A Golgi and electron- microscopic study in Macaeus rhesus. J Comp Neurol 141:283-312

Rakic P (1972) Mode of cell migration to the superficial layers of fetal monkey neocortex. J Comp Neurol 145:61-84

Rakic P (1975) Cell migration and neuronal ectopias in the brain. In : Birth defects : Original Articles Series 11 : 95-129

Rakic P (1981) Neuronal-glial interaction during brain develop- ment. Trends Neurosci 4:184-187

Salm AK, Smithson KG, Hatton GI (1983) Lactation-associated glial plasticity in the supraoptic nucleus of the rat. Soc Neurosci Abs 9:451

Schiebler TH, Leranth C, Zaborszky L, Bitsch P, Rutzel H (1978) On the glia of the median eminence. In: Brain-endocrine interaction III. Neural hormones and reproduction. 3rd Int Symp, Wfirzburg 1977, Karger, Basel, pp 46-56

Silver J, Sidman RL (1980) A mechanism for the guidance and topographic patterning of retinal ganglion cell axons. J Comp Neurol 189:101-111

Singer M, Nordlander RH, Egar M (1979) Axonat guidance during embryogenesis and regeneration in the spinal cord of the newt: the blueprint hypothesis of neuronal pathway patterning. J Comp Neurol 185 : 1-22

Tweedle CD (1983) Ultrastructural manifestations of increased hormone release in the neurohypophysis. In: Cross BA, Leng G (eds) The neurohypophysis: Structure, function and control, Prog Brain Res, Vol 60. Elsevier Science Publ BV, pp 259-272

Tweedle CD, Hatton GI (1980) Glial cell enclosure of neurosecretory endings in the neurohypophysis of the rat. Brain Res 192:555-559

Tweedle CD, Hatton GI (1982) Magnocellular neuropeptidergic terminals in neurohypophysis: rapid glial release of enclosed axons during parturition. Brain Res Bull 8 : 205-209

Uemura H, Nozaki M, Kobayashi H (1981) Luteinizing hormone releasing-hormone in the cerebrospinal fluid (CSF) of the rat; a technique for collection of CSF from the third ventricle. Neuroendocrinol Lett 3:159-163

Wittkowski W, Scheuer A (1974) Functional changes of the neuro- nal and glial elements at the surface of the external layer of the median eminence. Z Anat Entwicklungsgesch 143:255-262

Accepted March 12, 1985