| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Anatomy and Cell Biology, Columbia University, New York, New York 10032
Address all correspondence and requests for reprints to: Dr. Ann-Judith Silverman, Department of Anatomy and Cell Biology, Columbia University, New York, New York 10032.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine the site of origin of GnRH cells, embryos were exposed to tritiated thymidine and killed 5 h later. Most GnRH cells incorporated label in the olfactory epithelium; however, some autoradiographically labeled GnRH cells, possessing a neuronal morphology, were found in the olfactory nerve and the forebrain, suggesting that some GnRH neurons divide as they migrate.
A cumulative labeling method employing tritiated thymidine was used to examine the birth date of GnRH neurons. Postmitotic GnRH cells were first detected at stages 19–21. At stage 24, a peak in GnRH neurogenesis preceded the increase in GnRH neurons expressing their peptide at stage 26. After stage 24, there was a gradual addition of postmitotic cells to the population through stage 35. A pulse-chase paradigm indicated that birth date did not influence the final GnRH cell distribution. Injections at stage 29, when 10% of the GnRH neurons are born, generated double labeled cells in all locations where placode-derived GnRH neurons reside.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
GnRH neurons are among a growing class of migratory cells that differentiate in the olfactory placode and migrate to the central nervous system (CNS) (7, 8). In the chick, GnRH neurons differentiate in the olfactory epithelium (OE), migrate along fascicles of the olfactory nerve (ON) to reach the CNS (9, 10), and cross the cribriform plate at the medial margin of the olfactory bulb. They then course ventromedially into the forebrain (1). Within the CNS, the neurons first arch dorsally across the telencephalon, then turn to migrate ventrally and caudally to reach the septal, preoptic, and anterior hypothalamic areas (1).
To acquire quantitative data on the temporal and spatial patterns of the chick GnRH population, embryos were analyzed at individual stages, as the chick embryo may pass through several developmental stages per day, and the timing of in ovo development may vary among investigators due to dissimilar incubation temperatures and other environmental conditions. In the spatial dimension, no quantitative description of the regional distribution of GnRH neurons has yet been provided for this species at any age. Our observations were structured to answer four questions. 1) When and where do GnRH neurons first express their peptide? 2) When do GnRH neurons leave the olfactory epithelium and begin to migrate toward the brain? 3) When do GnRH neurons enter the telencephalon? 4) How many cells comprise the GnRH population at each time point, and how are they distributed among the OE, ON, and CNS?
Earlier in situ hybridization and immunocytochemical studies in which GnRH messenger RNA and GnRH peptide were first observed in the olfactory epithelium support the hypothesis that GnRH neurons originate in the olfactory placode (7, 8, 9, 10, 11). Olfactory placode ablation experiments result in the elimination of GnRH neurons on the ipsilateral side (12). These experiments, however, do not distinguish between mitotically active precursors and postmitotic GnRH cells, and thus do not eliminate the possibility that precursors may divide and differentiate outside of the olfactory placode. From preliminary, short survival tritiated thymidine (3HT) labeling experiments, we know that mitotically active, immature GnRH cells can be labeled with 3HT and detected with autoradiography (13). To examine GnRH cell proliferation more closely, we pulse-labeled embryos with 3HT and examined the distribution of radioactive GnRH neurons at short intervals thereafter to determine the site of origin of GnRH neurons.
In contrast to the GnRH system, most CNS cells are generated from the neural tube in specialized proliferative zones (14, 15, 16). Neuronal cell precursors divide actively within the ventricular and subventricular zones to generate their postmitotic progeny. The neurons then migrate to often distant cortical sites (14, 15, 16, 17, 18, 19). Spatio-temporal regulation of cell migration is a key feature in brain development. Classical studies using 3HT (14, 15, 16, 19) revealed that there is a close relationship between the time of origin of cells and their final location. Cerebral cortical layers display an "inside-out" neuronal migration pattern with respect to their time of cell origin. In contrast to the mammalian cortex, the chicken telencephalon develops in an "outside-in" pattern of neurogenesis, such that cells born early migrate to more lateral regions, whereas cells born later occupy more medial positions closer to the ventricles (17, 18).
The role of birth date in the final, complex distribution of GnRH cells is unknown. Unlike the cortex, the GnRH system is not a laminated structure. Rather, GnRH neurons are scattered topographically throughout the ventral forebrain, although they are often found as cell clusters (20). Likewise, neurons that ultimately reside in nonlaminated structures, such as specific hypothalamic nuclei, or in particular locations within the thalamus share a spatio-temporal origin in the ventricular neuroepithelium (21, 22, 23, 24).
Although much is known about GnRH neuronal differentiation, migration, function, and distribution (1, 2), little is known about the time of origin of GnRH neurons and the temporal relationship between their birth date and final location. In one previous study, 3HT autoradiography and GnRH immunocytochemistry revealed that GnRH neurons in the embryonic mouse are born between embryonic days 10–11 (E10–11) (11). The third aim of our study was to examine the time of origin of GnRH neurons and the potential role of birth date in GnRH neuronal distribution. Birth dates were determined using a modified cumulative labeling procedure with 3HT autoradiography and GnRH immunocytochemistry to determine the time of origin of GnRH neurons in the chick. Preliminary pulse-chase procedures were then employed to determine whether there a correlation between GnRH neuronal birth dates and their settling pattern.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Three, 3-week-old male broiler chicks [two from Arbor Acres (Blairsville, GA) one from Shaver (Shaver Poultry Breeding Farms, Cambridge, Ontario, Canada) provided by W. Kuenzel] were anesthetized with Choloropent (1.8 ml/kg), then perfused via the heart with heparinized saline and 4% paraformaldehyde solution in 0.1 M PB (pH 7.2). The heads were postfixed overnight in 4% paraformaldehyde. Before sectioning on a Vibratome (Dako Corp., Carpenteria, CA), the brains were removed from the skull, and the beaks were decalcified overnight in Decal (Omega Chemical Corp., Cold Spring, NY).
GnRH immunocytochemistry. Embryos for GnRH immunocytochemistry were killed at stages 17 (n = 3), 18 (n = 2), 19 (n = 2), 20 (n = 2), 21–23 (n = 5), 24 (n = 4), 25 (n = 2), 26 (n = 2), 27–28 (n = 5), 29 (n = 3), 32 (n = 3), 35 (n = 3), 42 (n = 3), 45 (n = 3), and posthatch p21 (n = 3; provided by W. Kuenzel). Embryos (or dissected brains and beaks) were embedded in 8% gelatin and cut in either the sagittal or horizontal plane at a thickness of 100 µm on a Vibratome (DSK Microslicer, Ted Pella). GnRH immunocytochemistry was performed on free floating sections from each of these embryos. Sections were washed in 0.1 M PB, pH 7.2, and treated in 0.5% H2O2 to remove endogenous peroxidase activity. All incubations longer than 2 h were performed at 4 C. Shorter incubations were performed at room temperature. The tissue was washed overnight in PB with 0.1% Triton X-100 (Sigma Chemical Co., St. Louis, MO) containing 3% normal horse serum, followed by a 5-day incubation with a monoclonal antibody, HU4 (1:750; provided by Dr. H. Urbanski) (26). We found that prolonged incubation (5 days or more) in primary antiserum greatly enhances the sensitivity of the procedure and results in labeling of greater numbers of GnRH neurons (27). HU4 recognizes the cleaved, mammalian form of GnRH (26). The antibody was diluted in PB plus 0.1% Triton X-100 with 1% normal horse serum. Bound antibody was detected using an overnight incubation in biotinylated horse antimouse IgG (1:200, Vector Laboratories, Inc., Burlingame, CA) followed by an avidin-biotin-horseradish peroxidase complex (Vector Laboratories, Inc.) for 2 h. The chromogen used was either 3,3'-diaminobenzidene (Sigma Chemical Co.), which yields a brown precipitate, or the Sligtly Grey substrate kit for peroxidase (Vector Laboratories, Inc.), which yields a blue-gray stain. The hydrogen peroxide was generated using glucose oxidase.
Di-I tracing of migratory cells. Embryonic chicks were prepared for Di-I injection during the stages when the olfactory placode is visible macroscopically (stages 17–19) (28). Three milliliters of albumin were removed from each egg with a sterile syringe. Windows were created to expose the embryos. Embryonic membranes were dissected from the face using tungsten needles to gain access to the olfactory placode. The vital dye, Di-I (Molecular Probes, Inc., Eugene, OR; 2.5 mg/ml) in 10% ethanol was injected onto the olfactory placode using a glass micropipette pulled to an od of 5–10 µm. After injection, the windows were sealed with clear tape, and the eggs were returned to the incubator. The embryos were then killed at stage 25 (n = 20). Di-I-labeled animals were cryoprotected in sucrose overnight at 4 C and frozen in isopentane cooled in dry ice. The embryos were sectioned serially on a cryostat (Bright/Hacker Institute, Inc., Fairfield, NJ) at 15–20 µm and thaw-mounted onto double subbed slides. The sections were coverslipped using GelMount (Biomeda, Foster City, CA), and Di-I-labeled cells were visualized using the rhodamine excitation and emission spectra.
Controls. The specificity of the monoclonal antibody against mammalian GnRH and chicken-I GnRH has been demonstrated previously (26). HU4 binds to only the mature bioactive form of GnRH and does not recognize the precursor or degraded form of the peptide. Thus, fragments of GnRH, such as those generated by enzymatic degradation of the molecule, are not bound by the antibody. HU4 does recognize the deamidated form of the decapeptide. HU4 demonstrates a high degree of cross-immunoactivity with chicken-I GnRH, but at approximately 41% the efficiency as mammalian GnRH (26). The mammalian and chicken-I GnRH decapeptides differ at position 8, where the mammalian form has a glycine and the chicken-I form has an arginine. The ventral forebrain GnRH cells express chicken-I (26). Chicken-II is expressed by the midbrain GnRH neurons, and these were not immunoreactive with the HU4 antibody (Mulrenin, E. M., personal observation).
Data analysis. Tissue sections were examined by light microscopy (Olympus Corp., New Hyde Park, NY) at x400 to identify immunocytochemically labeled GnRH neurons. By adjusting the focal plane, all GnRH neuronal nuclei throughout each tissue section were visualized and counted. Because the diameter of a GnRH neuron is 10 µm, and the cytoplasm is scant compared with the nucleus, no correction for double counting was used.
Unpaired t tests were used for statistical analysis of GnRH cell numbers. Statistical significance was assigned to P < 0.05.
An autoradiographic analysis of the site and time of origin and
settling pattern of GnRH neurons
This study is based upon a series of chick embryos injected with
3HT. Mitotically active cells were exposed to
3HT, which is incorporated into the replicating DNA. These
cells appear labeled in autoradiographs by the presence of silver
grains over their nuclei. Unlabeled cells include those neurons that
are not in S phase, are postmitotic during 3HT exposure, or
neurons that have undergone multiple cell divisions and have thus
diluted the radiolabel to undetectable levels. Three different
3HT injection protocols (Exp II, III, and IV) were used to
examine chick GnRH neuron proliferation.
Exp II: site of origin (pulse-labeling method with short
survival)
3HT injections. Chicks at
developmental stages 23 (n = 3), 24 (n = 2), 25 (n = 5),
26 (n = 2), 27 (n = 3), 28 (n = 3), and 29 (n = 2)
were exposed for 5 h in ovo to 10 µCi (25 Ci/mmol)
3HT [10 µl in 90 µl Tyrode’s (Sigma Chemical Co.) buffer] and then killed. These stages span 2.0–2.5
days.
Histological procedures. Embryos were fixed in 4% paraformaldehyde. After fixation, two different tissue processing techniques were performed. Some embryos were cut on a cryostat (Hacker/Bright) to a thickness of 15 µm and processed subsequently for GnRH immunocytochemistry as described above in Exp I. Other brains were cut using a Vibratome (Dako Corp.) to a thickness of 100 µm and processed for GnRH immunocytochemistry as described for Exp I, and regions containing GnRH-positive cells (OE, ON, and telencephalon) were microdissected, osmicated for 1 h in 2% OsO4 (Electron Microscopy Sciences, Fort Washington, PA) in 0.9% NaCl2 and 1.5% K3Fe(CN)6, dehydrated in alcohols and propylene oxide, and embedded in Epon resin (Tousimis Research Corp., Rockville, MD). Selected Epon blocks were sectioned at 0.5 µm on an ultramicrotome (AO Reichert, Wien, Austria) and counterstained with toluidine blue (0.1 M borate, pH 11).
Autoradiographic procedures. Immunostained sections were dried thoroughly, dipped in NTB-2 nuclear track emulsion (Eastman Kodak Co., Rochester, NY; 44 C; diluted 1:1 with water), and exposed for 2–4 weeks in a dry atmosphere at 4 C. The emulsion-covered slides were developed in Dektol (Eastman Kodak Co.) for 3 min (16 C), followed by a wash in water for 30 sec (16 C). The slides were then immersed in Rapid Fix (Eastman Kodak Co.) for 5 min (16 C) and washed twice in water. The first wash was for 1 min (16 C); the second was for at least 5 min (16 C). Some sections were counterstained with cresyl violet. The slides were then dehydrated in graded alcohols, cleared in Histoclear (Fisher Scientific Corp., Fairlawn, NJ), coverslipped with Permount (Fisher Scientific Corp.), and viewed by light microscopy (Olympus Corp.; x400).
For Exp III and IV, immunocytochemical and autoradiographic procedures were identical to those described here.
Quantitative evaluation. All slides were coverslipped with Permount (Fisher Scientific Corp.) and viewed by light microscopy (Olympus Corp.; x400). The 15-µm cryostat sections were further analyzed by counting all GnRH neurons. The cells were categorized as labeled or unlabeled with 3HT, and the proportion (percentage) of labeled cells was then calculated using the following formula (19): % labeled GnRH cells = (number of labeled GnRH cells/total number of GnRH cells) x 100%.
Exp III: time of origin (modified cumulative labeling)
3HT injections. Windows were cut in
the egg shells at stage 17, and embryos were placed in the
incubator. On the experimental day, embryos at stages 19–21 (n =
6), 24 (n = 3), 27 (n = 3), 28 (n = 2), 29–30 (n =
2), and 35 (n = 2) were exposed to three 100-µl injections of 10
µCi 3HT (Amersham, Arlington Heights,
IL; 25 Ci/mmol) in Tyrode’s (Amersham) buffer, spaced
4 h apart,. After 3HT injections, the tape was
replaced, and the embryos were incubated for varying durations.
Histological procedures. Embryonic heads were fixed in 4% paraformaldehyde or Zamboni’s fixative, and serial sagittal sections (15 µm) were cut on a cryostat (Bright/Hacker Institute, Inc.). Sections were processed for GnRH immunocytochemistry as described in Exp I using either a monoclonal antiserum against the mammalian GnRH decapeptide, HU4 (26), or a polyclonal antiserum against the mammalian decapeptide, LR1 (gift from R. Benoit). Detection of bound antibody was accomplished using biotinylated horse antimouse or goat antirabbit IgG (1:200; Vector Laboratories, Inc.) overnight, followed by avidin-biotin-horseradish peroxidase complex (Vector Laboratories, Inc.) for 2 h. The chromogen used was 3,3'-diaminobenzidine (Sigma Chemical Co.). Immunocytochemistry preceded autoradiography.
Quantitative evaluation. A distinction between heavily and lightly labeled cells is not necessary with the cumulative labeling procedure (19). All GnRH neurons observed in the plane of the nucleus were counted and assigned to one of two groups: labeled or unlabeled. The proportion of unlabeled cells was then given using the following formula (23): % unlabeled GnRH cells = (number of unlabeled GnRH cells/total number of GnRH cells) x 100%.
When a progressive increase was observed in the number of unlabeled
cells from a minimal level as a result of delays in the onset of
3HT application, it can be concluded that the precursor
cells are producing postmitotic neurons (i.e. neurons are
being born). By analyzing the rate of increase in unlabeled neurons, we
can determine the proportion of neurons born (ceased incorporation of
tritiated thymidine) on, or between, particular stages or days in
development (see Table 1
and Fig. 6).
|
|
3HT injections. We exposed stage 29 embryonic chicks (n = 2) to two pulses of 10 µCi 3HT (10 µl 25 Ci/mmol in 90 µl Tyrode’s buffer) in ovo, spaced 6 h apart. These radioactive pulses were followed by one 100-µl injection of a 100-fold excess of cold thymidine (0.2 mM thymidine in Tyrode’s buffer; Amersham). Embryos were then killed at stage 42, when the GnRH population had reached its adult distribution.
Quantitative evaluation. All GnRH neurons that were observed in the plane of the nucleus were counted, mapped by location, and categorized as labeled or unlabeled with 3HT.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
summarizes our quantitative
and spatial analyses of GnRH neurons across developmental time. We
present below those key events that answer the questions posed.
Photographic illustrations are limited to a few significant events in
the GnRH migratory process. GnRH neurons differentiated and initiated
migration during the olfactory placode stages of development, with the
first faintly immunostained cells observed in the OE at stage 19 (Fig. 1). A few hours later, at stage 20, two or three cells in some, but not
all, embryos had exited the basal compartment of the OE (Fig. 1).
Studies in our laboratory demonstrate that these early GnRH neurons
migrate on the first axons of the ON (29). These axons reach the
telencephalic border by stage 19 (29). By stage 21, the olfactory pits
have invaginated, forming nasal grooves with a thickened epithelium
from which emerge small axonal fascicles that converge in the ON (Fig. 2A). GnRH neurons were found within the
small fascicles and along the entire nerve. They migrated as a
continuous stream of closely associated cells from stage 21 onward
(Fig. 2, A and B). Some GnRH neurons arrived at the junction of the ON
and the telencephalic vesicles by stage 21 (Fig. 2A), just 12 h
after the first cells exited the OE.
|
|
) and did
not immediately enter the CNS. Rather, they paused at this location for
eight stages (
3 days) until stage 29, when the cells entered the
brain (Fig. 2D). The presence of this compact cellular mass was
confirmed using in ovo Di-I application to the early
olfactory placode. At stage 25, Di-I-positive cells were seen in the ON
and in the cluster outside of the telencephalon, but not in the CNS
(Fig. 3).
|
and 2D). The GnRH cells in
the nerve tended to be bipolar and were found in linear arrays (Fig. 2, A–C); those in the CNS possessed tapered, sometimes multipolar,
neurites (Fig. 2D). The cells dispersed upon entering the CNS (Fig. 2D).
An examination of the spatial-temporal distribution of the GnRH
population revealed trends in migration that may be useful for further
elucidation of GnRH neurogenesis. We review the ontogeny of the total
GnRH population and the partitioning of GnRH neurons among three
anatomical regions comprising their migratory route below. These
anatomical partitions include: 1) nasal epithelium (olfactory and
respiratory epithelium), 2) ON, and 3) CNS (see Fig. 1 for the pattern
of expansion). The total GnRH population expanded continuously
throughout embryogenesis, with two peaks. The first significant
increase occurred between stages 25 and 26 (P =
0.0031). This increase was due to a rise in the number of GnRH cells in
the ON (P = 0.0002). The second significant expansion
in the total population occurred between stages 35–42
(P = 0.0234). This increase was largely due to an
increase in the CNS component of the population.
There were no statistically significant differences in nasal epithelium-resident GnRH cell numbers between contiguous stages of development. Thus, the number of GnRH neurons in the olfactory (and respiratory) (30) epithelium increased modestly in successive stages through stage 29. There were, however, statistical differences in GnRH cell numbers in the nasal epithelium between early developmental stages (olfactory placode and pit stages) and late embryonic stages (nasal epithelium with distinct sensory and respiratory domains). For example, the difference in cell number (2 vs. 670) between stages 19 and 29 gave P < 0.0323. Immunoreactive cells persisted in this location until stage 35. Between stages 35–42, the GnRH phenotype was lost from the epithelium.
GnRH neurons within the nerve exhibited a different pattern. They showed a 10-fold expansion between stages 25–26 (P = 0.0002), increasing to approximately 1000 cells. The size of the population remained relatively stable for the rest of in ovo development. GnRH neurons persisted in the ON in the posthatch chick in comparable numbers (mean of 700 cells). The presence of this neuronal subpopulation in the posthatch chick has not been noted previously (see Discussion).
Despite the significant increase in migratory GnRH cells at stage 26, they remained arrested at the border of the presumptive olfactory bulb until stage 29 (see above). The percentage of immunoreactive neurons entering the forebrain at stage 29 was low, but statistically significant (P = 0.0435) and expanded over the course of in ovo development. There was another substantial increase in the CNS population between stages 32–42 (P = 0.0371), which correlated with the increase in total cell number that occurred at stage 42. In the posthatch chick, the central population remained highly variable in number, reaching about 5400 of 6600 total cells in a 3-week-old chicken. These data indicate a potential for generating a much larger GnRH population than has been reported for any other animal.
The adult distribution as reported by Kuenzel and Blahser (20) for GnRH neurons was achieved by stage 35. These cells were widely distributed throughout the ventral forebrain, from the olfactory bulb to the septal/preoptic areas, and into the anterior hypothalamic region of the diencephalon. GnRH neurons extended axons toward the presumptive median eminence by stage 35 (data not shown).
In stage 42–45 embryos and in the 3-week-old chick, immunoreactive GnRH neurons were also found in the lateral anterior nucleus of the thalamus. Thalamic GnRH neurons were few in number (<30).
An autoradiographic analysis of the site and time of origin and the
settling pattern of GnRH neurons
We examined the site of origin of GnRH neurons by determining the
regions in which GnRH precursors were mitotically active
(i.e. became radiolabeled after a 5-h pulse of
3HT). The data were first analyzed as the percentage of
GnRH neurons that were radiolabeled, treating each compartment
separately (see Fig. 4). As expected,
most GnRH neurons were mitotically active while present in the OE (Fig. 4). At all ages examined, a higher percentage of the OE-resident GnRH
cells was mitotically active compared with ON- and CNS-resident GnRH
neurons. However, some radiolabeled cells were found in the nerve of
all stages examined (Fig. 4). In addition, in one embryo injected at
stage 28, 2% of the labeled GnRH cells were found in the septal region
of the telencephalon (Fig. 4).
|
). The entire epithelium was labeled
heavily, thus serving as an internal control for our experimental
design (Fig. 5). The majority of the non-GnRH, mitotically active cells
presumably represent the sensory neuron stem cell population that
proliferates extensively during embryogenesis (31).
|
and Fig. 6.
When embryos were exposed to 3HT at stages 19–21, most of
the GnRH precursors were still mitotically active (86%), and only a
small percentage had withdrawn from the mitotic cycle (14%). In
contrast, application of 3HT at stage 24 revealed that
approximately 57% of the population had been born. Of this 57%, about
14% had withdrawn before stages 19/20. An additional 42% withdrew
between stages 19 and 24. Therefore, 43% of GnRH precursors were still
mitotically active at stage 24. By the end of stage 28, about 75% of
the GnRH cells had been born, and by stages 29/30, about 90% of the
GnRH population had become postmitotic. The last series of embryos was
injected at stage 35 and fixed at stage 42. Approximately 3% of the
GnRH neurons were radiolabeled when examined at stage 42 and,
therefore, had been mitotically active at stage 35. Thus, GnRH
neurogenesis continues at a slow rate at stage 35.
To determine whether a GnRH neuron’s time of origin influences the
neuron’s final destination, 3HT pulse-chase experiments
were performed on stage 29 embryos. The birth-dating experiments
described above indicated that 10% of the population was mitotically
active at this stage. Embryos were killed at stage 42. There was no
apparent preferential localization of the radiolabeled GnRH neurons;
they were found in the ON and olfactory bulb, the parolfactory bulb
area, the septal/preoptic region, the nucleus accumbens, and the
anterior hypothalamus (Table 2). In all
regions containing GnRH neurons (with one exception) both labeled and
unlabeled cells were often clustered together, as illustrated in the
nucleus commissura pallii of the telencephalon (Fig. 7). Although there were unlabeled
GnRH-immunopositive cells in the nucleus lateralis anterior thalami,
there were no 3HT-labeled GnRH cells in this nucleus, which
is derived from the diencephalic neuroepithelium, not from the
olfactory placode (32).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our examination of the olfactory epithelium demonstrates that cells containing GnRH appear very early when the epithelium is still considered placodal and has only a few sensory axons projecting from it (29). In the other well studied species, the mouse, GnRH neurons in the epithelium appear between E10.75 (34) and E11.5 by GnRH immunocytochemistry (7, 8, 11) and in situ hybridization (11). At this time, the epithelium has divided into an OE and vomeronasal organ (VNO), and GnRH neurons are present within the medial olfactory pit as well as the VNO. Thus, in both species, GnRH neurons appear at approximately the same developmental stage with respect to their extracranial origin.
In both the chick and the mouse there is also a coincidence between the appearance and (apparent) disappearance of GnRH neurons in the OE and their translocation into the CNS. Our data show that the number of GnRH cells in the chick OE never exceeds a few hundred, even when much larger numbers appear in the nerve (or CNS), suggesting that there is a rapid transition from acquisition of peptide to exiting the epithelium. It is also possible that GnRH-immunoreactive cells, or their progenitors, are capable of dividing outside of the olfactory epithelium while migrating. Our pulse-chase data support this latter hypothesis (see below).
Up to 20% of the total GnRH population remains extracranial in the 3-week-old chick. This population has been noted previously by Murakami and co-workers (9). In the adult mouse, 10% of the GnRH population also persists in the VNN and terminal nerve (TN) (34). A TN has been identified in the chick (35), and it is possible that some of our cells reside in fascicles of this nerve. If so, then the posthatch peripheral GnRH neurons may perform an (unknown) function associated with the TN and its target sites within the CNS. In fish, the TN projects peripherally to the retina and centrally to the preoptic and supracommissural nuclei of the area ventralis of the forebrain and may thus coordinate olfactory and visual cues. Demski and colleagues (36) reported a potentially novel function of GnRH neurons in the preoptic area of the goldfish, in that injection of GnRH into the preoptic area of the fish resulted in the sustained release of sperm.
Many GnRH neurons reached the boundary of the ventral forebrain at stage 21 and paused for several stages (3 embryonic days) at the telencephalic border, suggesting that they are inhibited temporally by environmental cues. These cues may include changes in the cellular environment, such as an interplay among the various cellular adhesion molecules or the extracellular matrix at this junction.
At the border of the CNS, the olfactory nerves cross the foramina of the cribriform plate and travel a short distance within the subarachnoid space before entering the nerve fiber layer of the olfactory bulb. At this junction, there is an abrupt change in the glial cell type, from ensheathing glia (peripherally) to oligodendrocytes (centrally). The basal lamina of the olfactory nerve at this site is continuous with that of the glial limitans of the subpial astrocytes, and the nerve is surrounded by connective tissue (37). This continuity is achieved developmentally. If we postulate that GnRH migration into the brain is regulated by the environment at the nerve-brain border then GnRH neuronal entrance into the CNS may require molecular changes in extracellular matrix at the telencephalic border in the olfactory nerve itself.
Such changes are observed during olfactory neuronal axogenesis. For example, Gong and Shipley (38) and Treolar and colleagues (39) found laminin and heparan sulfate proteoglycan in the early developing olfactory pathway of mice and rats, respectively, and in the meninges surrounding the telencephalon. As olfactory neurites breached the telencephalic border, the laminin- and heparan sulfate proteoglycan-positive meninges broke down on the surface at the presumptive olfactory bulb; chondroitin sulfate proteoglycan (CSPG) expression was down-regulated in this region concomitantly. Thus, this regional degradation of the basal lamina of the telencephalon and the down-regulation of CSPG within the marginal zone may facilitate the passage of primary olfactory axons into the brain to form the presumptive nerve fiber layer of the olfactory bulb. A similar mechanism (an interplay of matrix molecules) may regulate GnRH neuronal migration into the brain.
Another potential player in nerve/telencephalic interactions is the product of the KAL gene associated with X-chromosome-linked Kallmann’s syndrome (40, 41). The KAL gene product is an extracellular glycoprotein with potential serine protease inhibitory and cell adhesion domains (40, 41). There is some evidence that the KAL gene product binds the heparan sulfate chains of proteoglycans (40, 41) and thus may interact with the molecules discussed above. Whether a deficit in KAL expression is the immediate causative factor inhibiting olfactory axon entrance into the CNS is not known. It is, however, postulated that the unsuccessful migration of GnRH neurons into the CNS is secondary to axonal failure (6).
It is also possible that the GnRH neuronal pause at the telencephalic border is a cell-autonomous event and that GnRH neurons require a period of time to acquire the intracellular machinery necessary to breach the telencephalic border. Intracellular mechanisms are employed by cerebellar granule cells that demonstrate a stop-start pattern of migration. In studies of dissociated cerebellar cells by Edmondson and Hatten (42), granule cells migrated for 30–120 min along Bergman glia, then paused for 1 or 2 h before resuming migration. This saltatory migration pattern was confirmed in real-time examination of Di-I-labeled granule cells in acute cerebellar slice preparations and is thought to depend on the cyclic oscillation of intracellular Ca2+ levels (43). Those levels may, in turn, be regulated by multiple intrinsic and extrinsic mechanisms, including various voltage- and ligand-activated Ca2+ channels (43). We recognize that granule cells pause for relatively short time periods (1–2 h) compared with a longer GnRH pause (2.5 days). Regardless of the temporal discrepancies, similar cellular mechanisms may be employed by the GnRH system to regulate cellular motility.
At stage 29, GnRH neurons reinitiated migration and crossed into the
brain. Although GnRH cells did not enter the CNS until this time, other
migratory cells may have entered earlier. In embryos to which Di-I was
applied to the placode at stage 19, GnRH negative/Di-I-positive cells
were found in the brain as early as stage 25 (44). The selective
passage of these GnRH-negative cells into the brain before GnRH neurons
suggests that cell migration into the brain may show some degree of
cellular specificity. There are several candidate olfactory
placode-derived migratory cells that follow the same GnRH migratory
route. These include cells positive for carnosine and olfactory marker
protein (45), somatostatin (46), acetylcholinesterase (47),
neuropeptide Y (30), and 
-aminobutyric acid (48). Because a
small number of nonpost mitotic GnRH cells enter the brain, we cannot
rule out the possibility that some of the Di-I-positive cells entering
the brain at stage 25 could be nondifferentiated GnRH neurons.
Once within the CNS, GnRH neurons find their final locations by stage 35. The mechanisms by which individual cells choose their destinations are unknown. In the developing cerebral cortex, migratory stop signals are thought to be provided (at least in part) by the afferent fibers entering the target location or by the neuronal cell populations that had already reached their final positions (49). The birth date of a neuron has also been shown to play a role in determining the cell’s final location (14, 15, 16, 17).
Our studies also quantitate the number of GnRH neurons present in the lateral anterior nucleus of the thalamus. Thalamic GnRH neurons have been reported previously by Norgren and Gao (32) and are thought to be of diencephalic origin, expressing a peptide similar, perhaps identical, to mammalian GnRH.
An autoradiographic analysis of the site and time of origin and
settling pattern of GnRH neurons
Site of origin. The three different experiments using DNA
synthesis provide us with considerable detail on the genesis of GnRH
neurons. Using a short pulse protocol, we have shown that the majority
of GnRH precursors were mitotically active (i.e.
radiolabeled) while in the olfactory epithelium. These results support
the hypothesis proposed by olfactory placode ablation and
transplantation studies that the olfactory placode is the site of
origin of GnRH neurons (12, 50). However, after a 5-h pulse of
3HT, radiolabeled GnRH cells were also found in the
olfactory nerve and forebrain, suggesting that these cells may have
been progressing through S phase (and possibly M phase) along their
migratory route while making the GnRH peptide. This phenomenon of
preneuroblasts dividing as they migrate has been described by other
investigators. Hicks and D’Amato (51) found 3HT-labeled
cells outside of the proliferative ventricular zone, en route to the
isocortex, in the embryonic rat 5 h after a 3HT
injection. Menezes and co-workers (52) and Rousselot and colleagues
(53) later also identified a population of anterior subventricular
zone-derived neuroblasts that replicate while migrating to the
olfactory bulb in the postnatal mouse and rat.
If autoradiographically labeled GnRH cells found outside of the epithelium (in the olfactory nerve and brain) were not dividing as they migrated, then they must have finished dividing in the epithelium and moved rapidly into the nerve. Thus, some labeled GnRH neurons may pass through G2 and complete M in the epithelium, and migrate out of the epithelium immediately thereafter at rates rapid enough to reach these sites (ON and CNS) within 5 h. A determination of GnRH migratory rates is necessary before we can distinguish between these two hypotheses.
It is unclear whether the radiolabeled, immunoreactive GnRH cells observed 5 h after the 3HT pulse were in S and concurrently synthesizing GnRH or if they made GnRH before (re)entering the cell cycle. Equally, the remaining cells that were immunoreactive, but did not incorporate thymidine, could have been postmitotic at the time of 3HT exposure or could have been "waiting" to reenter the cell cycle. This is particularly true for the GnRH-only cells observed between stages 23–26. The possibility that preneuroblasts with neuronal characteristics can divide (as well as migrate) has been demonstrated for some neural crest cells by Rothman and colleagues (54). In this case, migrating crest cells, exhibiting many aspects of the adrenergic phenotype, including catecholamine synthesis, storage, and reuptake, were mitotically active. Our data suggest, but do not prove conclusively, that some GnRH-containing cells do not wait to become postmitotic to acquire aspects of their neuronal phenotype.
Neither the birth-dating experiments nor those employing the short pulse speak to the issue of cell cycle dynamics for the GnRH neuron or to the size of the progenitor population. We found that application of a brief pulse (seconds) of a replication deficient retrovirus at stages 17–21 labeled only about 1% of the GnRH neurons observed at stage 35. Retrovirus can only be incorporated into the genome of cells in S at the time of infection. In contrast to GnRH neurons, the olfactory epithelial cells were heavily labeled (13). This suggests that at the time when the GnRH progenitors/preneuroblasts are expanding rapidly (as measured by 3HT application), they do not incorporate the viral genome efficiently. This could be due to an asynchrony in neuroblast division or to a relatively short S phase. As the labeling is very high in the OE, it is clear that GnRH neurons have a very different proliferation pattern from the olfactory epithelial cells (31). This observation reemphasizes the possibility, first raised by elAmraoui and Dubois (55), that the GnRH progenitors vary in a substantive way from those progenitors that give rise to the olfactory sensory neuron.
Time of origin of GnRH neurons
Using the cumulative labeling paradigm for analysis of birth date,
it is clear that chick GnRH neurons are born over an extended period of
time, encompassing many developmental stages. However, the patterns of
generation over this period (E3 to at least E9) vary. Some cells
withdraw from the mitotic cycle very early (stage 19–20), before the
appearance of GnRH immunoreactivity, whereas other GnRH cells are born
as late as stage 35. Between stages 19/20 and stage 24, a peak in GnRH
neurogenesis results in the production of approximately half of the
GnRH population and their withdrawal from the cell cycle, but it is not
until 12 h later, at stage 26, that there is a corresponding
increase in the number of immunoreactive cells from 100 to nearly 1500
(13). These data suggest that the majority of GnRH neurons born during
the initial phase withdraw from the mitotic cycle before producing
their distinctive peptide.
The generation of postmitotic cells in the chick occurs over several stage (stages 19–35). A neurogenic peak produces approximately half of the cells; the rest are added to the population sequentially over many days of in ovo development. Indeed, 6.5% of GnRH neurons are created between stages 29/30 and 35 (E9). Although this is but a small fraction of the whole, it should be noted that as few as 1–16 cells can produce a functional hypothalamic-pituitary-gonadal axis in the hpg mouse (56). There is a major difference in GnRH neurogenesis between chick and mouse. In the mouse, GnRH neurons are born between E10 and E11 (11). In contrast, chick GnRH neurogenesis occurs over an extended period of embryogenesis.
Chick and mouse may differ in how the GnRH neurons are generated, because neurogenesis is occurring in different tissues. In rodents, the first cells appear in the presumptive VNO (7, 8), whereas in the chick, which lacks a VNO, they are first seen in the OE (9, 10) and, to a smaller degree, in the respiratory epithelium (30). The OE and the VNO are two anatomically and functionally distinct sensory organs. The olfactory sensory neurons bind odorant molecules and transduce the chemical signal into physiological odorant recognition (31). Vomeronasal neurons are activated by pheromones, which elicit innate social and sexual behaviors (57). The odorant receptors of the olfactory epithelium and pheromone receptors of the VNO share no homology and thus may have evolved independently (58, 59).
Relationship between time of origin and settling pattern of GnRH
neurons
Altman and Bayer (60) proposed a neurogenetic hypothesis to
explain the relationship among time and site of origin and settling
pattern of neurons in the developing CNS. These researchers proposed
that the adult nervous system is, to some extent, predetermined in the
neuroepithelium, and the directional gradients that cells form result
from a nonrandom distribution of the cells according to their age. This
hypothesis has been demonstrated in many neuronal systems using
3HT autoradiography as outlined in the introduction.
To test this hypothesis in the GnRH system, we examined a subpopulation of GnRH neurons (10% of the total population) that was generated at stage 29. Using a 3HT pulse and a cold chase, we found that the labeled GnRH neurons migrated to diverse regions of the embryo, including the olfactory nerve, the rostral and caudal telencephalon, and the anterior hypothalamic area of the diencephalon. These labeled GnRH cells were side by side with unlabeled GnRH neurons (except in the nucleus lateralis anterior thalami, where only unlabeled GnRH neurons were observed). Thus, the final destinations for GnRH neurons generated at stage 29 include all regions in which olfactory placode-derived GnRH neurons are found in the adult (20). Experiments are now in progress to examine other time points during GnRH neurogenesis.
The organization of GnRH neurons in the embryo may depend on a more complex pattern of neurogenesis. The data obtained from the 3HT experiments illustrate the migratory patterns of cell populations born at the same time, not those of individual cells. The settling pattern of GnRH neurons may be generated by a complex gradient and/or regional differences that are not obvious with our methods. We also do not know whether individual cells in the CNS stop migrating for a while and then move again later. Using retroviral technology, Cepko and co-workers (61) and Sanes and colleagues (62) demonstrated that a clone of cells can undergo widespread dispersion. These researchers used retroviral libraries comprising DNA tags as lineage markers to trace the cell lineage and migration of single cells and clonally related cells in a population. They found that clones of cells in the rat cerebral cortex (61), optic tectum, forebrain, and spinal cord (62) migrate initially radially, followed by widespread dispersion into several complex patterns. Similarly, an individual GnRH precursor cell may produce clonal progeny that are destined for many regions of the brain, thereby ensuring the dispersal of the GnRH population to all target destinations.
The present studies demonstrated that the chick GnRH population, with its interesting migratory behavior and its continued expansion in ovo and posthatch, is a useful model system for analyzing the mechanism underlying neurogenesis. The fact that the questions that arise from this can be addressed both in ovo (Drapkin, P., and A.-J. Silverman, in preparation) and in vitro (50) make it particularly appealing.
We have presented a quantitative, immunocytochemical study of the temporal and spatial dimensions of GnRH differentiation and migration during chicken embryogenesis. Our most significant finding is that GnRH neurons pause for several stages on the olfactory nerve at the telencephalic border before entering the CNS. Using Hamburger and Hamilton’s (25) staging system, we have defined when and where GnRH neurons begin to differentiate, the time of onset of migration into the olfactory nerve and brain, and the distribution and number of cells at several developmental stages. The second most significant observation is that the GnRH population increases over time. Our studies show that the GnRH population expands throughout embryogenesis, with major increases occurring in two peaks. In some animals, a third population peak occurs postnatally.
Using 3HT autoradiography and GnRH immunocytochemistry, we
have also determined the site and time of origin of GnRH neurons and
the relationship between GnRH neuronal birth dates and their final
settling pattern. Most GnRH precursors divide in the olfactory
epithelium. However, some autoradiographically labeled GnRH cells are
found in the olfactory nerve and the brain after a 5-h pulse of
3HT, suggesting that some cells divide as they migrate. Our
studies have revealed that GnRH neurons are born throughout
embryogenesis, with a large percentage (50%) of the cells
withdrawing from the mitotic cycle between stages 21 and 24. We have
provided preliminary evidence that GnRH neurons generated at stage 29
migrate to all regions in which olfactory placode-derived GnRH neurons
are found in the adult.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received November 25, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-aminobutyric
acid and gonadotropin-releasing hormone during neuronal migration
through the olfactory system. Endocrinology 137:5415–5420[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  S. Bhattacharyya and M. Bronner-Fraser Competence, specification and commitment to an olfactory placode fate Development, December 15, 2008; 135(24): 4165 - 4177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Giacobini, A. S. Kopin, P. M. Beart, L. D. Mercer, A. Fasolo, and S. Wray Cholecystokinin Modulates Migration of Gonadotropin-Releasing Hormone-1 Neurons J. Neurosci., May 19, 2004; 24(20): 4737 - 4748. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. M. Erhardt, E. A. Fradinger, L. A. Cervini, J. E. Rivier, and N. M. Sherwood Early Expression of Pituitary Adenylate Cyclase-Activating Polypeptide and Activation of its Receptor in Chick Neuroblasts Endocrinology, April 1, 2001; 142(4): 1616 - 1625. [Abstract] [Full Text]
|
|
|
|
|
|
S. X. Simonian and A. E. Herbison Differing, Spatially Restricted Roles of Ionotropic Glutamate Receptors in Regulating the Migration of GnRH Neurons during Embryogenesis J. Neurosci., February 1, 2001; 21(3): 934 - 943. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. F. Striedter and B. P. Keefer Cell Migration and Aggregation in the Developing Telencephalon: Pulse-Labeling Chick Embryos with Bromodeoxyuridine J. Neurosci., November 1, 2000; 20(21): 8021 - 8030. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Maggi, F. Pimpinelli, L. Molteni, M. Milani, L. Martini, and F. Piva Immortalized Luteinizing Hormone-Releasing Hormone Neurons Show a Different Migratory Activity in Vitro Endocrinology, June 1, 2000; 141(6): 2105 - 2112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. S. Parhar, T. Soga, and Y. Sakuma Thyroid Hormone and Estrogen Regulate Brain Region-Specific Messenger Ribonucleic Acids Encoding Three Gonadotropin-Releasing Hormone Genes in Sexually Immature Male Fish, Oreochromis niloticus Endocrinology, May 1, 2000; 141(5): 1618 - 1626. [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 |
