| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Division of Endocrinology and Metabolism, Mount Sinai School of Medicine, New York, New York 10029
Address all correspondence and requests for reprints to: Dr. Marie J. Gibson, 19530 Mammoth Drive, Bend, Oregon 97702. E-mail: jpgmjg{at}bendcable.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cografts of fetal pituitaries with POA into the third ventricle of hypogonadal mice result in GnRH fibers growing into the gland, in addition to median eminence innervation (11). The finding that grafted GnRH neurons innervate the median eminence of hypophysectomized hosts indicates that the anterior pituitary is not an essential source of attractive factor(s). Indeed, in cografts of POA and fetal medial basal hypothalamus (MBH), a mini median eminence was formed within the graft, with robust GnRH innervation and GnRH axons terminating near fenestrated capillaries. To further study this apparent targeting, we used coexplants of fetal or neonatal POA with MBH, cerebellum, or spinal cord on collagen- and laminin-coated membranes in insert chambers (12). More and longer GnRH fibers grew out upon the membrane from the POA explant only toward MBH coexplants as early as 4 days and was maintained through 10 days of culture. These results were not due to trophic influences on the survival of GnRH neurons or fibers, as there were no differences in total numbers of GnRH cells within the cultures regardless of the nature of the coexplants, nor were there differences in the total numbers of GnRH axons growing out upon the membranes. Rather, the findings strongly suggest that diffusable chemoattractive signals for GnRH neurons are derived from the MBH.
The present studies were designed to establish that the MBH is a source of a soluble chemoattractant(s) to GnRH outgrowth while excluding possible cellular factors. We used presoaked heparin-coated acrylic beads anchored to the membrane in a drop of collagen and laminin to present putative targeting substances to the POA explant. First, we evaluated conditioned medium derived from MBH explants. As the fenestrated capillaries in the median eminence are the ultimate targets in the MBH, we also used beads presoaked in conditioned medium from mouse brain microvascular endothelial cells. In an initial effort to determine specific candidates for chemoattraction to the GnRH axons, we tested the efficacy of basic fibroblast growth factor (bFGF), an important component of both endothelial cells and ventricular tanycytes. The generality of the effect of MBH medium on other neuroendocrine cells was assessed by determining its effect on galanin fiber outgrowth.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clusters of heparin-coated acrylic beads (Sigma) were incubated in a microfuge tube containing the test substance or control medium for 1–2 h. These beads have been used to present growth factors in studies of limb development in chick embryos (13). Approximately 3–5 µl of the collagen/laminin preparation were placed on the insert membrane, and a small cluster of the presoaked beads was placed in the drop and allowed to dry. Chambers were labeled sequentially, and experimental and control inserts were randomly placed. Codes for which chambers contained which types of beads were kept separately to permit blind analyses. POA tissue explants were then prepared as described below and placed immediately on the collagen- and laminin-coated surface of the insert chamber membrane at a small distance from the coated beads. All explants were kept moist by the addition of 10–20 µl DM above the membrane. The chambers were then placed in wells of 12- well plates containing 1 ml of the same medium. Cultures were incubated for 1 week at 37 C in 5% CO2 at 100% humidity. In all experiments, the medium was not changed, and care was taken to leave the cultures undisturbed until the time of fixation. The institutional animal care and use committee of the Mount Sinai School of Medicine approved all procedures.
Tissue collection
Newborn pups (C3H/HeHx101H) were collected less than 24 h
after delivery (P1) and anesthetized by placement in culture dishes on
ice before decapitation. The head was rinsed with a solution of 70%
ethanol before dissection to assure sterility and was placed on a
sterile stage under a binocular dissection microscope. The skull was
removed, and the brain was inverted on the stage so that the ventral
surface of the brain faced up. The POA was dissected under sterile
conditions as described previously (14). When preparing MBH cultures
for MBH-conditioned medium, the tissue including the median eminence
was dissected as a midline strip from the ventral surface of the brain
caudal to the POA. Tissue explants were placed immediately on the
collagen- and laminin-coated surface of the insert chamber
membrane.
Preparation of conditioned medium and specific factors
For MBH-conditioned medium (MBH-CM), three or four MBH explants
obtained from P1 pups were placed on the coated surface of each
membrane and covered with 100 µl DM. Each insert was placed in a well
of a sterile 12-well plate containing 1 ml DM/well. Cultures were
incubated for 1 week at 37 C in 5% CO2 at 100%
humidity. Medium from the surface of an insert was aspirated and placed
in a microfuge tube. The amount obtained was approximately 25 µl. An
additional 100 µl DM were obtained from the well, used to rinse the
membrane, and added to the microfuge tube. Each collection consisted of
medium from three to six chambers. After collection, the
MBH-conditioned medium was freed of any cellular matter, sterilized by
filtering through a 0.2-µm pore size filter, and placed in a sterile
tube before freezing at -20 C. DM was used as control. For endothelial
cell-conditioned medium, mouse brain microvascular endothelial cells
(MVEC; Cell Applications, Inc., San Diego, CA) were subcultured with
MVEC growth medium, trypsinized, and replated with the DM described
above, containing no serum, hormones, or antibiotics. After 3 days of
incubation, the endothelial cell-conditioned medium was collected,
freed of cellular material, and sterilized with a 0.2-µm pore size
filter, and aliquots were frozen at -20 C. DM was used as a control. A
product of endothelial cells, bFGF, has been implicated in regulating
GnRH secretion. Beads were soaked in an aliquot containing 1 µg bFGF
(recombinant human bFGF, Oncogene Research Products,
Cambridge, MA) or DM.
Immunocytochemistry
After 7 days of incubation, the cultures were fixed by immersion
in 4% paraformaldehyde for 3 h. All fixed cultures were kept at 4
C in phosphate buffer (0.1 M; pH 7.3) with 0.1% sodium
azide (Sigma) until they were processed.
Immunocytochemistry was performed directly on membranes after cutting
them out of the chamber. To detect GnRH neurons and processes, SW1
antiserum (15) was used at a dilution of 1:2500 in PB containing 0.1%
Triton X-100 (Sigma), 1% normal goat serum (Life Technologies, Inc.) or 3% normal donkey serum (NDS)
(Jackson ImmunoResearch Laboratories, Inc., West Grove,
PA) for 4 days at 4 C. We previously showed that preabsorption of the
antiserum with 10 µg GnRH peptide eliminated reaction product in
cells and fibers. The biotinylated antirabbit secondary antiserum
(1:200 made in donkeys; Jackson ImmunoResearch Laboratories, Inc.) in the same diluent was applied overnight for increased
penetration of the explants. Explants were then incubated in an
avidin-biotin solution (Vector Laboratories, Inc.,
Burlingame, CA) conjugated with horseradish peroxidase for 2 h.
The chromogen was 1 mg/ml 3,3'-diaminobenzidine (Sigma)
and the H2O2 was generated
by the oxidation of glucose by glucose oxidase (Sigma). To
determine whether MBH-conditioned medium affected the outgrowth of
another neuroendocrine peptide, other POA cultures were processed for
galanin immunoreactivity using an antigalanin antibody raised in
rabbits (Peninsula Laboratories, Inc., Belmont, CA;
1:4000). No immunoreactivity was detected when the antibody (1 ml) was
preincubated with 200 µg synthetic galanin. The antibody was diluted
as for the SW-1 used for GnRH as described above, and the cultures were
similarly incubated, processed, and analyzed.
Immunocytochemistry was performed to detect the presence of bFGF in tanycytes in the mouse brain. Male mice deeply anesthetized with chloral hydrate (700 mg/kg, ip) were perfused with saline followed by 4% paraformaldehyde. Brains were removed and postfixed overnight, and coronal sections, 50 µm thick, were obtained with a Vibratome (Vibratome, Inc., St. Louis, MO). The sections were washed and incubated for 72 h with a dilute anti-bFGF antibody (raised in rabbits; Chemicon International, Inc., Temecula, CA) in 0.1 M PBS (1:1000) containing 1.5% NDS and 0.3% Triton X). After washing, bFGF was visualized using a mixture of 3,3'-diaminobenzidine and 0.005% H2O2 after incubation of the sections with 0.5% donkey antirabbit biotinylated antibody (Jackson ImmunoResearch Laboratories, Inc.) for 3 h and avidin biotin peroxidase solution (Vector Laboratories, Inc.) for 1 h. After washing, the sections were then incubated for 72 h in a dilute (1:1000 in 0.1 M PBS containing 1.5% NDS and 0.3% Triton X) antivimentin antibody (mouse monoclonal V2258, Sigma). In control sections for either bFGF or vimentin, there was no staining if the primary antibody was deleted. The sections were then washed, mounted onto gelatin-coated slides, coverslipped using Gelmount mounting medium (Biomedia, Foster City, CA), and analyzed under a light and fluorescent microscope (AX70, Olympus Corp., Melville, NY).
Analyses
Cultures were not considered for quantitative analysis if the
explants fused with the beads or if there were less than 50
GnRH-immunoreactive (GnRH-ir) cell bodies in the POA. There are
proposed spatial limits on guidance of 1 mm for target-derived
diffusable substances and 1 cm for a substrate-bound gradient (16). We
did not exclude cultures on the basis of distance between the culture
and the cluster of beads, however, as the possibility remains that a
diffusable substance may interact with elements in the substrate. GnRH
axons extending on the surface of the membrane from the POA explant
were counted on the side facing the cluster of beads and on the side
away from the beads. It should be noted that fiber length only refers
to distance from the edge of the explant and does not account for the
variable locations of GnRH neurons within the explants.
The data were analyzed using statistical software (GBSTAT, Dynamic Microsystems, Inc.). All values were expressed as the mean ± SEM. When homogeneity of variances was not achieved, the appropriate transformation of the data were applied. The number of GnRH axons on the membranes toward and away from the beads was compared between treatment and control cultures using two-way ANOVA. When P < 0.05, Fisher’s least significant difference was used for post-hoc comparisons. The maximum extent reached by GnRH axonal outgrowth from the POA explant was also assessed. The three longest extending GnRH axons from the POA explant were measured in each sector with the help of a micrometer grid inserted in the eye piece of an Olympus Corp. BH-2 light microscope as described by Lumsden and Davies (17). The measurements were defined as follows. The zero reference bar of the grid was placed tangentially to the interface between the POA explant and membrane. The maximum distance reached on the membrane by the growth cone of the three farthest extending GnRH axons was measured regardless of the winding path of the axon. These three lengths were averaged for the sides toward and away from the beads. Due to the nature of the data, the maximum axonal extent was analyzed using nonparametric tests; Kruskal-Wallis ANOVA was used to compare multiple groups. Comparisons between the sides toward and away were analyzed using Wilcoxon signed rank test. Analysis for trophic effects on GnRH cell survival was performed by counting the number of GnRH cell bodies in the POA in the presence of treated or control beads. Comparing total GnRH axonal outgrowth between the groups using a t test assessed possible trophic effects on GnRH axonal outgrowth. The cultures were all coded before analysis so that the type of bead was unknown at the time of counting (e.g. they were counted blind).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). As before, the explants tended to
flatten during the culture period, resulting in varying distances
between the edge of the culture and the cluster of beads (Table 1). GnRH fibers could also be followed as
they exited the explants onto the surface of the insert chamber
membrane (Fig. 1B), where growth cones were often discernible. The
cluster of beads (Fig. 1C) was easily visualized, and occasionally
fibers approached the cluster (Fig. 1D).
|
|
, upper panel).
All but 1 of the 18 explants had more GnRH fibers toward the MBH-CM
beads (
2 = 14.2; P < 0.01).
GnRH fibers also extended farther upon the membrane toward the MBH-CM
beads than on the side away (P < 0.01). In 15 of the
18 cultures, the longest GnRH fibers were toward the MBH-CM beads
(2 = 8; P < 0.01). In
contrast, similar numbers and lengths of GnRH fibers grew on each side
of the control POA explants (n = 12) that were cocultured with DM
(see Fig. 2). In just 4 of the 12 control cultures there were more
fibers toward the beads; in 6 the longest fibers were toward, and in 6
they were away. There was no difference in the distance between the POA
and the cluster of beads between the 2 groups (Table 1
). Conditioned
medium derived from the MBH did not have a trophic influence on GnRH
survival or axonal outgrowth. The numbers of GnRH-ir cells within the
POA explants were similar whether cocultured with beads presoaked in
MBH-CM or in DM, as were the total numbers of GnRH-ir fibers growing
out upon the membrane (Table 1).
|
Effect of MVEC-CM
More GnRH-ir fibers grew toward the MVEC-CM beads than grew either
on the side away from the MVEC-CM beads or on the side away from
the control beads (P < 0.05; Fig. 3, upper panel). The
difference in outgrowth was not significant within the control cultures
(P > 0.10). A large proportion (17 of 24;
2 = 10.4; P < 0.01) of these
POA cultures had more fibers growing toward the MVEC-CM beads. In
contrast, the proportion of the control cultures with more GnRH fibers
growing toward the beads was just 8 of 18 (not significant). In neither
group were there differences in the length of fiber outgrowth (Fig. 3, lower panel).
|
, upper panel). Of 15
explants cocultured with bFGF-soaked beads, 13 had more GnRH-ir fibers
on the side growing toward the beads (2 =8.1;
P < 0.01). The mean length of the 3 longest fibers was
significantly less on the side away from the bFGF beads than on the
sides toward the bFGF beads or the control beads (P <
0.05; Fig. 4, lower panel). The proportion (11 of 15) of the
explants with longer fibers toward the bFGF beads was not significant.
The number and length of GnRH-ir fibers on the membrane were similar on
each side of the control cultures. The tendency for the total number of
GnRH fibers to be lower in cultures in the presence of bFGF did not
reach significance (P > 0.06; Table 1).
|
). In previous work (18) we visualized a
larger population of tanycytes, perhaps due to the different antibody
for vimentin employed here. From the pattern of bFGF labeling it is
likely that bFGF is present in only a portion of the tanycytes in the
median eminence.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The MBH region contains the median eminence, the ultimate target of GnRH axons in normal development and also of GnRH fibers emerging from third ventricular POA grafts in hypogonadal mice. In both cases GnRH fibers terminate upon the fenestrated capillaries in the median eminence. We hypothesized that the endothelial cells in these capillaries contain a chemoattractive factor(s) for GnRH axons. The results with conditioned medium from the cultured mouse brain MVEC supported this hypothesis, but were not impressive. The mouse brain endothelial cells may have been an imperfect model for the endothelial cells in the capillaries of the median eminence, as the latter are fenestrated, in contrast to those in most neural vasculature. Recent findings indicate considerable heterogeneity among endothelial cells from different sources as well as changes in characteristics due to culture conditions (19).
In earlier work we discussed the role of tanycytes in GnRH neuronal guidance at the median eminence and noted that we did not always see close associations (18). At that time we speculated that although tanycytes may provide a permissive substrate for GnRH axonal extension, they do not fully explain the precise targeting seen, but perhaps tanycytes play a role not just in physical guidance by possible adhesion factors, but also by the presence of guidance factors such as bFGF in at least a portion of them. Kozlowski and Coates (20) described the close contact of GnRH fibers with basal processes of tanycytes, often within ependymal canaliculi formed by adjacent ependymal cells and also containing unidentified axons. We showed that when GnRH axons grow out upon the membrane from POA explants, they are not usually associated with tanycytes, but always travel with growth associated protein (GAP-43)-labeled axons (12). Within the MBH coexplants, however, GnRH axons projected to a region resembling the median eminence, as visualized by vimentin-ir tanycytes. Tanycytes are also implicated in regeneration of monoaminergic axons in the MBH (21).
The use of heparin-coated beads may have enhanced the effectiveness of bFGF, as it is a member of the family of heparin-binding growth factors. Widely distributed during embryonic development (22) and in neuronal and glial cells in the adult rodent brain (23, 24), bFGF is also present in the ependymal cells that line the third ventricle (25) and in endothelial cells (26, 27). Many structures in the rat hypothalamus contain the messenger RNA for the bFGF receptor, FGFR1 (24). This receptor contains a cell adhesion molecule domain that is implicated in neurite outgrowth (reviewed in Ref. 28). FGF activity also involves the extracellular signaling molecules heparan sulfate proteoglycans, and there is the possibility that the activity may sometimes involve a ternary complex of heparan sulfate proteoglycan, FGF, and the FGF receptor (28). Nevertheless, the mechanism of bFGF release remains unclear, because it lacks a signaling domain.
Survival and neurite outgrowth is promoted by bFGF in GT1 cells, the immortalized GnRH neuronal cell line (29, 30). The growth factor is also involved in processing pro-GnRH to decapeptide in GT1 cells (31). FGFR-1 receptors are present on GT1–7 cells (32), whereas it is difficult to detect these receptors on GnRH neurons. As explants dissected on P1 were used in the present study, it is highly likely that most GnRH axons growing out upon the membrane are regenerating. As in our previous work with both E15 and P1 POA explants (12), growth cones were often evident (not shown). In primary cultures, fetal GnRH neurons also respond to bFGF with neurite outgrowth (33). The present study further demonstrates that such outgrowth may be directed by bFGF to the appropriate target for neuroendocrine activity.
Galanin is another neuropeptide implicated in neuroendocrine function (34). The majority of galanin cells that project to the median eminence in the rat are located in the arcuate nucleus, as determined by uptake of peripherally administered fluorogold, with most of the others found in the paraventricular nucleus (35). In the mouse, however, very few galanin-ir cells in the arcuate nucleus have projections to the median eminence (36), whereas a large proportion of galanin neurons in the nucleus circularis, supraoptic nucleus, and paraventricular nucleus have such projections. The POA explants undoubtedly include at least the nucleus circularis. The large numbers of galanin-ir fibers seen here support that a substantial population of galanin neurons was present in the POA explants. Although one may argue that many of these neurons do not usually target the median eminence, the lack of effect of the MBH-conditioned medium on galanin fiber outgrowth is in striking contrast to the effect on GnRH neurons. As discussed above, even GnRH neurons obtained from the accessory olfactory bulb will target the median eminence from intraventricular grafts (9). The fact that galanin is often colocalized in GnRH neurons in the mouse (37), as in other species (38, 39), dramatizes the difference in control of axonal targeting. The present studies did not address whether MBH-conditioned medium preferentially attracts the axons of the subset of galanin neurons that colocalize GnRH. As they constitute a fraction of the total galanin neurons, it seems that even if the GnRH-containing fibers were attracted, it would be unlikely to significantly affect the total galanin-ir outgrowth measured here.
In our studies with hypogonadal mice we have shown that there is enormous redundancy in the reproductive system. The most remarkable example is the fact that a hypogonadal mouse may conceive and bear young with only one detectable GnRH neuron in the third ventricular graft as long as there is GnRH axonal innervation of the median eminence (6). Thus, although the work reported here indicates a role for bFGF in directing the outgrowth of GnRH projections, it is likely not unique. Consistent with this conclusion is the observation that the bFGF knockout mouse is fertile and not grossly phenotypically different from wild-type animal despite differences in neocortical architecture that suggest a role in neurogenesis (23, 40, 41). In light of the powerful need for reproductive success in conservation of species, these redundancies are not surprising. Another consideration in interpreting the results of the present studies is that there is a precedent for growth factors to act in concert. The model used here should be valuable in identifying other chemoattractive substances that direct GnRH targeting.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received February 14, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, and
insulin-like growth factor-I on a hypothalamic gonadotropin-releasing
hormone neuronal cell line. J Neurosci Res 49:739–749[CrossRef][Medline]
in the
release of luteinizing hormone-releasing hormone from the developing
female hypothalamus. Proc Natl Acad Sci USA 87:9698–9702This article has been cited by other articles:
 |
|
 |
|
  S. A. Tobet and G. A. Schwarting Minireview: Recent Progress in Gonadotropin-Releasing Hormone Neuronal Migration Endocrinology, March 1, 2006; 147(3): 1159 - 1165. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Gill and P.-S. Tsai Expression of a Dominant Negative FGF Receptor in Developing GNRH1 Neurons Disrupts Axon Outgrowth and Targeting to the Median Eminence Biol Reprod, March 1, 2006; 74(3): 463 - 472. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Pitteloud, J. S. Acierno Jr., A. U. Meysing, A. A. Dwyer, F. J. Hayes, and W. F. Crowley Jr. Reversible Kallmann Syndrome, Delayed Puberty, and Isolated Anosmia Occurring in a Single Family with a Mutation in the Fibroblast Growth Factor Receptor 1 Gene J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1317 - 1322. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. P. Bless, H. J. Walker, K. W. Yu, J. G. Knoll, S. M. Moenter, G. A. Schwarting, and S. A. Tobet Live View of Gonadotropin-Releasing Hormone Containing Neuron Migration Endocrinology, January 1, 2005; 146(1): 463 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Gill, S. M. Moenter, and P.-S. Tsai Developmental Regulation of Gonadotropin-Releasing Hormone Neurons by Fibroblast Growth Factor Signaling Endocrinology, August 1, 2004; 145(8): 3830 - 3839. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) or away from (
) the cluster of beads
presoaked in MBH-CM or control DM. Upper panel,
Significantly more GnRH fibers grew toward the MBH-CM beads than away
from them (P < 0.0l). Similar numbers of GnRH
fibers grew toward as away from the control beads. Lower
panel, The GnRH fibers growing toward the MBH-CM beads were
significantly longer (P < 0.01) than those growing
on the side away, whereas there was no significant difference in the
length of the fibers exiting the POA explants in the presence of
control beads.
