Reproduction in mammals is under the control of the hypothalamic neuropeptide gonadotropin hormone-releasing hormone-1 (GnRH-1). GnRH-1-secreting neurons originate during embryonic development in the nasal placode and migrate into the forebrain along olfactory nerves. Gradients of secreted molecules may play a role in this migratory process. In this context, hepatocyte growth factor (HGF) is a potential candidate, because it promotes cell motility in developing brain and has been shown previously to act as a motogen on immortalized GnRH-1 neurons (GN11). In this study, the role of HGF and its receptor Met during development of the GnRH-1 system was examined. GnRH-1 cells express Met during their migration and downregulate its expression once they complete this process. Tissue-type plasminogen activator (tPA), a known HGF activator, is also detected in migratory GnRH-1 neurons. Consistent with in vivo expression, HGF is present in nasal explants, and GnRH-1 neurons express Met. HGF-neutralizing antibody was applied to explants to examine the role of the endogenous growth factor. Migration of GnRH-1 cells and olfactory axon outgrowth were significantly reduced, in line with disruption of a guidance gradient. Exogenous application of HGF to explants increased the distance that GnRH-1 cells migrated, suggesting that HGF also acts as a motogen to GnRH-1 neurons. Functional experiments, performed on organotypic slice cultures, show that creation of an opposing HGF gradient inhibits GnRH-1 neuronal migration. Finally, tPA−/−:uPA−/− (urokinase-type plasminogen activator−/−) knock-out mice exhibit strong reduction of the GnRH-1 cell population. Together, these data indicate that HGF signaling via Met receptor influences the development of GnRH-1.
Gonadotropin hormone-releasing hormone-1 (GnRH-1) regulates anterior pituitary gonadotropes and, as such, is essential for reproduction. GnRH-1-secreting neurons originate from the nasal placode (Wray, 2002) during embryonic development and migrate to the hypothalamus apposed to olfactory-vomeronasal nerves (Schwanzel-Fukuda et al., 1989; Wray et al., 1989). In humans, several monogenic disorders leading to isolated hypogonadotropic hypogonadism (IHH) are caused by disruption of GnRH-1 neuronal ontogeny/migration (Gonzalez-Martinez et al., 2004). However, mutations in these genes do not account for many individuals exhibiting IHH. This suggests that the full repertoire of molecular cues regulating the GnRH-1 migratory process has not yet been identified.
Factors already shown to influence GnRH-1 neuron migration, either directly or indirectly via extension of olfactory axons (Wray, 2002; Wierman et al., 2004; Tobet and Schwarting, 2006), include neurotransmitters/neuropeptides (Fueshko et al., 1998; Bless et al., 2000; Simonian and Herbison, 2001; Pronina et al., 2003; Giacobini et al., 2004), surface molecules (Yoshida et al., 1999; Gamble et al., 2005), and growth factors (Cronin et al., 2004; Gill et al., 2004; Gill and Tsai, 2006). Guidance of the axonal/migratory pathway is also an important prerequisite for establishment of the adult-like GnRH-1 cell distribution (Wray, 2002), and classical chemoattractants [(e.g., netrin-1 and stromal cell-derived factor-1 (SDF-1)] or chemorepellents (e.g., reelin) are distributed in gradients along the GnRH-1 migratory route and participate in directing appropriate migration (Schwarting et al., 2001, 2004, 2006; Cariboni et al., 2005).
Hepatocyte growth factor (HGF) is a cytokine that, via its receptor Met, exhibits mitogenic, motogenic, and chemoattractive activities in neuronal (Ebens et al., 1996; Maina et al., 1997; Streit and Stern, 1997; Yamamoto et al., 1997; Caton et al., 2000; Ieraci et al., 2002; Gutierrez et al., 2004) and non-neuronal cells (Stella and Comoglio, 1999; Urbanek et al., 2005; Son et al., 2006). HGF and Met are widely distributed in developing brain (Jung et al., 1994; Thewke and Seeds, 1996; Achim et al., 1997; Thewke and Seeds, 1999; Korhonen et al., 2000); however, few studies address the function(s) of Met signaling during brain development. To date, HGF has been shown to have motogenic effects on migrating cortical neurons (Powell et al., 2001, 2003; Segarra et al., 2005). HGF is expressed in nasal embryonic mesenchyme with an increasing gradient toward the border between the nose and telencephalon (Sonnenberg et al., 1993; Thewke and Seeds, 1996). This pattern suggests that HGF/Met signaling might impact developmental events in the GnRH-1/olfactory system. In support of this, HGF exerts motogenic and chemotactic effects on the GN11 immortalized GnRH-1 cell line (Giacobini et al., 2002).
To determine the role of HGF in the developing GnRH-1/olfactory system, this study (1) characterized Met expression in nasal regions during the period of GnRH-1 neuronal migration, (2) perturbed HGF/Met signaling in two in vitro models (nasal explants and slice cultures) in which primary GnRH-1 neurons are maintained and cellular movement can be quantified, and (3) assessed the impact of the lack of HGF activators [plasminogen activators (PAs)] on the GnRH-1 neuronal population in PA knock-out (KO) mice.
Materials and Methods
Experiments were conducted in accordance with current European Union and Italian law, under authorization of the Italian Ministry of Health, number 66/99-A.
CD-1 embryos (Charles River Laboratories, Milan, Italy) were harvested at embryonic day 11.5 (E11.5), E12.5, E14.5, and E17.5 (plug day, E0.5) and used for RNA isolation, immediately frozen and stored (−80°C) until laser-capture microscopy, or postfixed [overnight; 4% paraformaldehyde (PFA) in 0.1 m phosphate buffer, pH 7.4] and cryoprotected and then frozen and stored (−80°C) until processing for immunocytochemistry. Tissue-type PA−/− (tPA−/−):urokinase-type PA−/− (uPA−/−)-deficient mice and wild-type (WT) background control mice (C57B16/129sv) were provided by Prof. P. Carmeliet [Center for Transgene Technology and Gene Therapy, Flanders Interuniversity Institute for Biotechnology, University of Leuven, Leuven, Belgium)]. CD-1 postnatal day 10 (PN10) mice and adult knock-out and WT animals were anesthetized with an intraperitoneal injection of ketamine (200 mg/kg) and perfused with 4% paraformaldehyde. The brains were dissected and postfixed in the same fixative overnight at 4°C, cryoprotected in sucrose solutions, and then frozen and stored (−80°C) until processing for immunohistochemistry.
Nasal regions were cultured as described previously (Fueshko and Wray, 1994). Briefly, embryos were obtained from timed pregnant animals in accordance with National Institutes of Health (NIH)/National Institute of Neurological Disorders and Stroke guidelines and Animal Care and Use Committee approval and with current European Union and Italian law. Nasal pits of E11.5 staged NIH-Swiss embryos were isolated under aseptic conditions in Gey's balanced salt solution (Invitrogen Grand Island, NY) enriched with glucose (Sigma-Aldrich, St. Louis, MO). Nasal explants were adhered onto coverslips by a plasma (Cocalico Biologicals, Reamstown, PA)/thrombin (Sigma-Aldrich) clot. The explants were maintained in defined serum-free medium (SFM) (Fueshko and Wray, 1994) at 37°C with 5% CO2. From culture day 3 to day 6, fresh medium containing fluorodeoxyuridine (8 × 10−5 m; Sigma-Aldrich) was given to inhibit proliferation of dividing olfactory neurons and non-neuronal explant tissue. The medium was changed to fresh SFM twice a week.
All primers were designed from published GenBank sequences and screened using BLAST (basic local alignment search tool) to ensure specificity of binding. Primers were pretested on brain cDNA and thereafter used throughout the described protocols at a concentration of 250 nm. Amplified products were run on a 1.5% agarose gel.
Reverse transcription-PCR analysis
Total RNA was isolated from noses and brains obtained from E11.5 mice using RNA STAT-60 (Tel-Test, Friendswood, TX) following the manufacturer's protocol. Briefly, the tissue was homogenized (1 ml of RNA STAT-60 per 50–100 mg of tissue), chloroform was added (0.2 ml/ml homogenate), and the mixture was spun. To the aqueous layer, isopropanol was added (0.5 ml) to precipitate RNA. RNA pellet was washed (75% ethanol), air dried, and resuspended (DEPC-treated water). Total RNA from adult mouse brain served as positive control tissue. For the reverse transcription (RT)-PCR, 0.5 μg of each sample was used. First-strand cDNA was synthesized using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen) following the manufacturer's instructions. PCR was performed using 4 μl of cDNA and the appropriate oligonucleotides in 30 μl PCRs using standard reaction buffer [(in mm) 10 Tris-HCl, pH 8.3, 50 KCl, and 1.5 MgCl2], 0.8 mm deoxynucleotide triphosphate (Invitrogen) and 0.025 U/μl REDTaq DNA polymerase (Sigma-Aldrich). The following primers were used: 5′-GGGACTGCAGCAGCAAAGC-3′ and 5′-GTCTGAGCATCTAGAGTTTCC-3′ for c-met amplification (Chan et al., 1988). For HGF, 5′-GGGGAATGAGAAATGCAGTCAG-3′ and 5′-CCTGTATCCATGGATGCTTC-3′ were used (Tashiro et al., 1990). The number of cycles and the annealing temperature used for each primer pair were as follows: 25 cycles and 59°C for c-met; 30 cycles and 55°C for HGF. No products were amplified in water or brain RNA not reverse transcribed.
Laser capture microdissection and RT-PCR on tissue-specific regions
Laser capture microdissection (LCM) permits cells to be isolated (“captured”) from tissue sections for molecular analyses. In this study, olfactory epithelium (OE), vomeronasal organ (VNO) epithelium, and lower jaw were captured from E14.5 and E17.5 mouse frozen sections (see Fig. 1 C,D) using a PALM LCM system (Zeiss, Thornwood, NY). The laser-microdissected tissues were popped into a sterile Microfuge cap containing 1 μl of 0.1% Triton X-100 and subsequently centrifuged for 1 min at 7500 × g (maximum) to relocate material to the bottom of a sterile tube. Prime RNase inhibitor (7 μl diluted 1:100 in DEPC-treated water; Eppendorf, Hamburg, Germany) was added. Captured tissue was used to synthesize first-strand cDNA using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen) following the manufacturer's instructions. Controls without reverse transcriptase were performed to demonstrate the absence of contaminating genomic DNA. Brain total RNA was also reverse transcribed and used as a positive control.
PCR was performed for βIII-tubulin (a general neuronal marker), early B-cell factor 2 (EBF-2) [an olfactory transcription factor and thus marker of olfactory/vomeronasal receptor neurons (Wang et al., 1997)], c-met, and HGF at 40 cycles on a thermocycler (30 s denaturation at 94°C, 30 s annealing at 55–65°C, and 2 min elongation at 72°C). PCR primer pairs were as follows: βIII-tubulin forward primer, 5′-GAGGACAGAGCCAAGTGGAC-3′; βIII-tubulin reverse primer, 5′-CAGGGCCAAGACAAGCAG-3′; EBF-2 forward primer, 5′-TGCAGTAGTT-GCTAACAGTGG-3′; EBF-2 reverse primer, 5′-TTTCCAATGCTAG-AAGCCTAAC-3′.
Cell isolation and PCR analysis
Nasal explants were washed twice with 1× PBS (without Mg+ or Ca2+) and placed in 2 ml of the same solution. GnRH-1-like neurons were identified by their bipolar morphology, association with outgrowing axons, and location within the explant (see Fig. 4 B). At two time points [4.5 and 28 d in vitro (div)], single GnRH-1 cells (n = 5 for each in vitro stage) were isolated from nasal explants using a micropipette (see Fig. 4 A–C) controlled by a micromanipulator (Narishige, Tokyo, Japan) connected to an inverted microscope (IX51; Olympus Optical, New Hyde Park, NY), cDNA was produced, and PCR amplification was performed as described previously (Kramer et al., 2000; Giacobini et al., 2004). Briefly, a single cell was lysed and reverse transcribed [AMV (avian myeloblastosis virus) and MMLV (Moloney murine leukemia virus)-reverse transcriptases; 37°C for 15 min; 65°C for 10 min] using an oligo-dT primer (50 OD/ml; pd(T)19–24). The cDNA was end labeled with terminal transferase (37°C for 15 min; 65°C for 10 min). Subsequent PCR amplification was performed using AL1 primers [ATT GGA TCC AGG CCG CTC TGG ACA AAA TAT GAA TTC (T)24] (Dulac and Axel, 1995) for 25 cycles in a DNA Thermal Cycler (94°C for 1 min, 42°C for 2 min, and 72°C for 6 min, with 10 s of extension time at each cycle; PerkinElmer, Wellesley, MA). After the first 25 cycles, fresh Taq was added, and 25 more cycles of PCR were performed minus the 10 s extensions. The resulting product was phenol-chloroform extracted and then ethanol precipitated, and an aliquot was run on a 1.5% agarose gel. Total brain RNA (1 μg) served as a positive control. All cDNA pools were initially screened for GnRH-1 (correct cell phenotype), β-tubulin, and L19 (two housekeeping genes, microtubule and ribosomal) using PCR. All cells used in this study were positive for all three transcripts. Primers sequences used were as follows: GnRH-1, 5′-GCTAGGCAGACAGAAACTTGC-3′ and 5′-GCATCTACATCTTCTTCTGCC-3′; β-tubulin, described above; and L19, 5′-CCTGAAGGTCAAAGGGAATGTGTTC-3′ and 5′-GGACAGAGTCTTGATGATCTCCTCC-3′. Each reaction mixture was generated as described above, and 2 μl of each primer and 1 μl of template cDNA were added. The PCR program was as follows: 10 min at 94°C prerun; 30 s at 94°C, 30 s at 55°C or 65°C (depending on primers), and 2 min at 72°C, for 35 cycles; and 10 min at 72°C postrun. The same PCR profile was used for subsequent screening with the following primers: tPA forward primer (5′-AAGTTTGCACTGGGGACAAG-3′), tPA reverse primer (5′-TCCCAAGAGTTGAGGAGTGTG-3′), uPA forward primer (5′-GTCTTCCATGTGATGCTCCA-3′), and uPA reverse primers (5′-ACCCAGTGAGGATTGGATGA-3′). Specific bands were observed in total E17.5 brain lanes, whereas no bands were seen in water lanes.
Primary antisera used were against GnRH-1 [SW-1, rabbit (Rb) polyclonal (Wray et al., 1988), kindly provided by Dr. S. Wray; LR-5, Rb polyclonal, kindly provided by Dr. R. Benoit, Montreal General Hospital, Montreal, Quebec, Canada; SMI41, mouse monoclonal antibody (Sternberger Monoclonals, Baltimore, MD)], HGF (#AF294-NA, goat polyclonal; R & D Systems, Minneapolis, MN) Met (#SP260 and #H-190, rabbit polyclonal, and #B-2, mouse monoclonal IgG; Santa Cruz Biotechnology, Santa Cruz, CA), peripherin (#AB1530, rabbit polyclonal; Millipore, Billerica, MA), neural cell adhesion molecule (NCAM; #C9672, mouse monoclonal IgG; Sigma-Aldrich), and tPA (#ASMTPA, rabbit polyclonal; Molecular Innovations, Southfield, MI).
Mouse tissue sections or nasal explants were immunocytochemically stained as described previously (Fueshko and Wray, 1994; Wray et al., 1994). Mouse embryos and postnatal and adult brains were cryosectioned respectively at 16 μm (embryos) and free-floating at 30 μm (postnatal or adult brains). These sections and explants were fixed with 4% formaldehyde for 1 h before immunocytochemistry. Briefly, sections or nasal explants were washed in PBS, incubated in 10% NGS/0.3% Triton X-100 (NGS/Tx-100; 1 h), washed several times in PBS, and placed in primary antibody (overnight at 4°C). The next day, tissues were washed in PBS, incubated in biotinylated secondary antibody [1 h; 1:500 in PBS/0.3% Triton X-100; goat anti-rabbit biotinylated (GAR-Bt; Vector Laboratories, Burlingame, CA); goat anti-mouse biotinylated (GAM-Bt; Millipore)], and processed using a standard avidin–biotin–horseradish peroxidase/3′, 3-diaminobenzidine (DAB) protocol. For double immunoperoxidase staining, the chromogen for the first antigen–antibody complex was DAB [brown precipitate (Kramer and Wray, 2000)], whereas the chromogen for the second antigen–antibody complex was SG substrate (blue precipitate; Vector Laboratories). Primary antisera dilutions were as follows: anti-GnRH-1 (SW-1; 1:3000), anti-peripherin (1:2000), anti-Met (#SP260 and # H-190; 1:200). For double-immunofluorescence experiments, primary antisera were diluted as follows: anti-GnRH-1 (SW-1, 1:1000; SMI41, 1:3000), anti-NCAM (1:60), anti-HGF (1:10), anti-Met (#SP260 and # H-190; 1:100), and anti-tPA (1:500). Sections or nasal explants were incubated overnight (4°C) in a mixture of primary antibodies diluted in NGS/Tx-100 and visualized using Alexa Fluor 488 and Alexa Fluor 568 conjugated secondary antibodies (1:500; Invitrogen). Anti-Met and anti-HGF antisera were incubated for two nights at 4°C.
When two polyclonal primary antibodies were used (anti-Met/SW-1 and anti-tPA/SW-1), staining of the first antigen–antibody complex was performed with goat anti-rabbit Alexa Fluor 488 (1:500; Invitrogen) secondary antibody. This step was followed by a blocking reaction with an anti-rabbit Fab fragment (Jackson ImmunoResearch Laboratories, West Grove, PA) (Giacobini et al., 2004) for 1 h (RT), followed by PBS washes, fixation (4% formalin for 30 min), and more PBS washes before application of the second primary antibody, which was visualized with conjugated fluorescent goat anti-rabbit cyanine 3 (Cy3; 1:800; Jackson ImmunoResearch Laboratories).
E12.5 slice cultures were fixed with 4% PFA for 1 h and processed for immunohistochemistry. To detect GnRH-1 immunoreactivity, slices were incubated at 4°C for five nights with LR-5 antibody diluted 1:4000 in PBS containing 10% NGS and 1% Tx-100. For secondary antibody processing, slices were washed several times with PBS for 1 h before incubation overnight at 4°C with goat anti-rabbit Cy3 (1:500; Jackson ImmunoResearch Laboratories).
Specimen were mounted in DABCO (1,4-diazabicyclo[2.2.2]octane; Sigma-Aldrich) and observed with a laser-scanning Olympus Fluoview confocal system (Olympus Optical).
To determine the function of HGF in the developing olfactory/GnRH-1 systems, pharmacological perturbations were performed, and olfactory axon outgrowth and GnRH-1 cell migration were quantified. Explants in experimental groups were maintained in SFM containing either a blocking HGF antibody (5 μg/ml; #AF294-NA; R & D Systems) or 25 ng/ml human recombinant HGF (Sigma-Aldrich). Drug concentrations were based on data from previous migrational studies (Powell et al., 2001; Giacobini et al., 2002). Nasal explants were treated at 3 div with pharmacological agents for 72 h. Control explants were maintained in SFM that was changed, as in the treatment groups, at 3 and 6 div. At 7 div, explants were processed for double-label immunocytochemistry for GnRH-1/peripherin (see above), and then the GnRH-1 cell migration as well as the maximum olfactory axon outgrowth were quantified as described below. Treatments performed at 1 div (for 72 h) did not induce any significant change in the parameters examined compared with control cultures (data not shown).
Quantification of GnRH-1 cell number, migration, and olfactory fiber outgrowth.
For each explant, the number of GnRH-1-immunopositive cells was counted on the main tissue mass, as well as in the periphery of the explant (Fueshko et al., 1998; Giacobini et al., 2004). The main tissue mass contained the nasal pit/olfactory epithelial region, surrounding mesenchyme, and nasal midline cartilage (see Fig. 3, schematic). The periphery refers to the area surrounding the main tissue mass into which cells had spread and/or migrated. Data are presented as mean ± SEM. Quantification of GnRH-1 cell migration and olfactory axon outgrowth was performed digitizing the stained explant. Images were taken under an Olympus IX50 inverted microscope (Olympus Optical) equipped with a CCD camera CoolSNAP-Pro (Media Cybernetics, Silver Spring) and images edited with Image-Pro Plus software (Media Cybernetics). For cell migrational measurements, a caliber with a series of concentric arcs separated by a uniform distance (200 μm) was overlaid on the digitized image. The total number of cells in each zone of the periphery of the explant was recorded via computer-assisted analysis (Image-Pro Plus software; Media Cybernetics) (see Fig. 6). In general, under all treatment conditions, the number of GnRH-1 cells decreases as a function of distance from the main tissue mass. GnRH-1 cell migration was calculated as the distance from the main tissue mass edge to the outer sector of the periphery.
For fiber outgrowth measurements, the distance from the border of the explant at which the majority of peripherin-positive fibers ended was recorded; this method was chosen because the complex nature of the fiber network prevented quantification of individual fiber lengths. A mean distance for fiber outgrowth was obtained for each treatment group, and values were reported as the mean ± SEM. All experiments used explants generated by different individuals on multiple culture dates.
Statistical analysis of fiber outgrowth was performed with a one-way ANOVA followed by Fisher's least significant difference (LSD) post hoc analysis (p < 0.05) using the statistical software SPSS 12.0 (SPSS, Chicago, IL). A mean total GnRH-1 cell number (inside the explant and/or in the periphery) was obtained for each treatment group and analyzed using a one-way ANOVA. These values were taken as an indication of GnRH-1 cell survival. Data for cell movement were compared for SFM and experimental groups by constructing contingency tables and applying the χ2 test for independence. This nonparametric analysis was chosen because zonal distance (200 μm grouping) was used instead of continuous measurements, and the number of observations per treatment (culture number) was not identical. A stringent p value of 0.001 was chosen for significance in these experiments.
To obtain HGF clones in which expression of murine HGF can be specifically induced by doxycycline (Dox), we adopted the tetracycline (Tet)-Off technology described by Gossen and Bujard (1992). The mouse HGF cDNA was inserted in the bidirectional pBI response plasmid containing enhanced green fluorescent protein (EGFP) as a reporter gene (Baron et al., 1995), which makes it possible to trace cells expressing the transgene by EGFP imaging. The plasmid was introduced into stable Madin-Darby canine kidney (MDCK) Tet-Off cells, and clones resistant to blasticidin were selected and expanded. The HGF cDNA construct was amplified from murine HGF plasmid (a kind gift from Dr. W. J. LaRochelle) using the Platinum Pfx polymerase (Invitrogen) and the following oligonucleotides (Sigma-Aldrich): 5′-TTGCACGCGTCCACCAT-GATGTGGGGGACCAAAC-3′ and 5′-TTACACGCGTGTTAACTTACTTTCCAAGTCGGTTCATCTCTATGTCTGTATACAACTTGTATGTCAAAA-3′. The obtained cDNA encoded the full-length HGF sequence, flanked with MluI sites and fused to a Kozak consensus ribosome binding site at the N terminus and to oligonucleotides encoding the 11 aa vesicular stomatitis virus glycoprotein G (VSVG) tag at the C terminus. The fidelity of the HGF insert was verified by sequencing (MWG Biotech, Ebersberg, Germany). The insert was then ligated into the MluI site of pBI-EGFP (Clontech, Mountain View, CA) to generate the HGF-TRE-EGFP responder plasmid. MDCK cells expressing the tetracycline transactivator (tTA) under cytomegalovirus promoter (MDCK-Tet-Off cell line; Clontech) were transfected with pBI-HGF-TRE-EGFP plasmid and a blasticidin selection plasmid. Exponentially growing MDCK-tTA cells were seeded 24 h before DNA transfer on 10 cm tissue culture dishes and transfected using Lipofectamine 2000 (Invitrogen). Cells were selectively grown in growth media containing 5 μg/ml blasticidin (Sigma-Aldrich) for 2 weeks. Different resistant clones were picked by selective trypsinization and checked for inducible expression of EGFP reporter gene and HGF by in vivo imaging of EGFP and scatter assay of conditioned medium, respectively (see Fig. 8). Clone n.1, which expresses high levels of EGFP- and HGF-transfected protein, was selected for this study.
Tet-Off cells were grown in monolayer at 37°C in 5% CO2, in DMEM (Invitrogen) containing 4500 mg of glucose, 1 mm sodium pyruvate, 2 mm glutamine, 100 μg/ml streptomycin, and 100 U/ml penicillin, and supplemented with 10% fetal bovine serum (FBS; Invitrogen). To turn off EGFP and HGF expression, Dox (1 μg/ml) was added to the culture medium. Cells within six passages were used throughout the experiments.
For Western blotting analysis, equal amounts of proteins (100 μg) were boiled in sample buffer (33% bromphenol blue, 33% β-mercaptoethanol, and 33% glycerol) and subjected to 7% SDS-PAGE. Proteins were blotted onto Hybond-C Extra membrane (GE Healthcare, Piscataway, NJ). Filters were probed with specific primary antibodies: 1:500 anti-mouse met #B-2 (Santa Cruz Biotechnology), 1:500 HGF goat antiserum (R & D Systems), 1:1000 P5D4 Mab (monoclonal antibody) mouse antiserum raised against the 11 aa C terminus of VSVG (Crepaldi et al., 1997) to detect HGF protein in lysates from MDCK cells expressing tagged HGF. In the latter case, total extracts were run under nonreducing conditions. Immunoblots were developed with an enhanced chemiluminescence kit, ECL (GE Healthcare).
Cell scattering assay
A classic scatter assay, using Met-expressing MDCK epithelial cells (Stoker et al., 1987; Montesano et al., 1991; Powell et al., 2001), was used to determine whether culture medium conditioned (CM) by nasal explants or Tet-Off cells contained functional HGF. MDCK cells were cultured in the same medium used for transfected cells, supplemented with 5% FBS (Invitrogen). MDCK cells (8000) were plated onto glass coverslips. The following day, discrete colonies were formed and then treated with known concentrations of HGF or CM (diluted 1:1 in fresh culture medium) collected from nasal explants at 3 div (medium conditioned for 3 d) or from Tet-Off cells cultured in the absence or presence of Dox (medium conditioned for 5–7 d). In a subset of experiments, a blocking-function HGF antibody was added (5 μg/ml). Twenty-four hours later, the MDCK cells were washed in PBS, fixed with 4% formaldehyde, and stained either by the nuclear cell dye 4,6-diamidino-2-phenylindole (DAPI) or by crystal violet. Images were taken under an Olympus IX50 inverted microscope (Olympus Optical) equipped with a CCD camera CoolSNAP-Pro (Media Cybernetics). Quantitative analysis of the scatter response was performed on digitized images that were overlaid on circles with a diameter of 80 μm (see Fig. 5 E). MDCK cells were counted within each circle superimposed on areas in which cells were detectable. In general, the number of cells within these counting frames decreases as a function of cell scatter. Values were reported as the mean ± SEM. ANOVA followed by Fisher's LSD post hoc analysis was used to compare groups (p < 0.001).
Tet-Off cell aggregates
Tet-Off cells were collected by trypsinization, resuspended in 20 μl of growth-factor free Matrigel (BD Biosciences, San Jose, CA) diluted 1:1 with the culture medium and seeded in 20 μl drops of this solution (200,000 cells for both cell lines, with or without Dox) on the lid of a culture dish. The lid was then turned upside down and incubated at 37°C for 10–20 min. As the droplets of cell aggregates were set, they were cut into four pieces (each one containing ∼50,000 cells) with a sterile blade.
Embryonic slice cultures
Timed pregnant CD-1 mice (Charles River Laboratories) were harvested at E12.5 to generate whole-head organotypic slice cultures following the procedures described previously (Tobet et al., 1996; Bless et al., 2000). Briefly, embryonic heads were embedded in 8% low-gelling-temperature agarose (type VIIa; Sigma-Aldrich), and parasagittal sections were cut at 300 μm using a vibratome and placed into Petri dishes containing ice-cold dissection medium (Leibovitz's L-15; pH 7.4; Invitrogen). These slices were moved carefully to avoid any torsion, stretch, or compression trauma, which may compromise the migration of GnRH-1 neurons in vitro.
E12.5 organotypic slices went through all of the steps described above until the point of plating. At this point, the tissue was fixed with 4% PFA and stained for GnRH-1 as described above.
All slices that were subjected to functional treatments were maintained in culture for 24 h. Organotypic slices were plated onto 30 mm Millicell inserts (Millipore) coated with a thin layer of growth factor-free Matrigel (BD Biosciences). The inserts were placed into dishes containing 2 ml of culture medium. This medium was partially removed from the wells, such that only a thin layer of liquid remained covering each slice. The slices were maintained in a humidified incubator (37°C) for 1 d. The slice culture medium consisted of Neurobasal medium (Invitrogen) supplemented with B27 supplement (Invitrogen), 0.5 mm glutamine (Invitrogen), and 25 μg/ml gentamycin (Invitrogen).
Aggregates of Tet-Off cells (50,000 cells) expressing EGFP–HGF (with or without Dox) were placed at the rostral tip of the nasal region at 0 div. After 24 h, organotypic cultures were fixed with 4% PFA and stained for GnRH-1 as described above. Slices were used for the quantitative analyses if they contained at least 250 GnRH-1-immunoreactive (GnRH-1-ir) neurons and if the connection between the nasal compartment and the brain appeared intact.
Quantitative analysis of GnRH-1 neurons was performed as a function of location with GnRH-1 cells assigned to one of two regions (nasal region and CNS). Total number of cells was calculated for each slice and combined to give group means ± SEM. Given that the total number of GnRH-1 neurons per slice did not change between treatment groups (see Results), GnRH-1 cell distribution is presented as the average percentage of labeled cells located in the nose or in the CNS under different treatment conditions. Where a significant overall one-way ANOVA was found (p < 0.05), post hoc comparisons using Fisher's LSD test were performed to further clarify significant differences between individual treatment groups (p < 0.05).
Quantitative analysis of GnRH-1 neurons in tPA−/−:uPA−/− mutant animals
Serial sagittal sections (30 μm; four series) from tPA−/−:uPA−/− and WT mice (n = 3 for each group) were cut and labeled for GnRH-1 as described above (GnRH-1 immunoreactivity visualized using DAB substrate). Total numbers of GnRH-1 cells were calculated in each brain and combined to give group means ± SEM. Data for GnRH-1 cell number between WT and knock-out animals were compared by one-way ANOVA followed by a Fisher's LSD post hoc test. ANOVA data were considered significantly different if p < 0.05.
HGF/Met expression in the developing nasal regions
To identify whether Met and HGF were expressed prenatally in nasal regions, nose tissue was removed at E11.5, when GnRH-1 neurons are in the presumptive VNO, and RT-PCR experiments were performed (Fig. 1 A,B). Both c-met and HGF transcripts were detected in all samples (E11.5 head, E11.5 nose, and adult brain) but water. Western blot analysis was used to document protein expression in embryonic nose tissues (Fig. 1 B, right). HGF is initially biosynthesized and secreted in a biologically inactive single-chain form (pro-HGF; ∼100 kDa) and is subsequently activated by specific serine proteases into an α-chain (69 kDa) and a β-chain (34 kDa) form containing a total of five glycosylation sites (Nakamura et al., 1989). E12 nose tissues and E16 head extracts (positive control) showed distinct bands corresponding to the α-chain (Fig. 1 B, right). Protein bands of α-chain were not single, showing that these proteins were glycosylated heterogeneously. The top band corresponds to the glycosylated form (69 kDa), whereas the bottom band represents the nonglycosylated α-chain (53 kDa). E12 nose protein extracts contained a barely detectable band of inactive single-chain HGF (100 kDa), indicating that most HGF in these tissues is the activated form. A 145 kDa Met-immunoreactive band was also evident in the same protein extracts (Fig. 1 B, right). These results indicate that at E12, both active HGF and its receptor are expressed in the nasal compartment.
To determine the expression of HGF and c-met transcripts in more defined regions of the nasal compartment, LCM was used. Single punches from olfactory epithelium (OE) and VNO were removed at E14.5, and RT-PCR experiments were performed (Fig. 1 C–E). This embryonic age corresponds to a stage of robust axonal outgrowth from the OE to the developing olfactory bulb. After reverse transcription of the mRNA, PCR with specific primers for β-tubulin (positive control) and the olfactory marker Olf/EBF-2 was performed (Fig. 1 E). A 158 bp band corresponding to β-tubulin (data not shown) and a 165 bp band corresponding to EBF-2 product were detected in both OE and VNO, supporting the olfactory nature of the laser-captured tissues (Wang et al., 1997). Products for c-met were found in the OE as well as in the VNO section, with a stronger expression in the latter tissue (Fig. 1 E). HGF transcript was not detected in the OE and VNO regions, whereas a specific band of correct size was found in the control lane (E17.5 whole-embryo extracts) (Fig. 1 E). These LCM RT-PCR data are in agreement with previous in situ hybridization studies showing c-met and HGF mRNAs in the developing murine OE and in the surrounding nasal mesenchyme, respectively (Thewke and Seeds, 1996).
To determine whether Met protein was expressed by the developing olfactory axons, double-label immunofluorescence was performed for Met and NCAM, a marker of the olfactory/vomeronasal system (Calof and Chikaraishi, 1989; Miragall et al., 1989). Met and NCAM expressions overlapped on fibers emerging from the VNO at E14.5, as shown by single confocal planes (Fig. 1 F,G, inset), and were coexpressed in olfactory/vomeronasal axon bundles from the nasal tract to the medial surface of the forebrain throughout the analyzed stages (E12.5–E17.5). Because of low signal-to-noise levels in brain, we were unable to detect specific immunoreactivity for Met along the caudal nerve that GnRH-1 cells follow into the ventral forebrain (Yoshida et al., 1995).
Migrating GnRH-1 neurons express Met
Immunohistochemistry indicated Met protein expression in the presumptive VNO epithelium as well as in cells migrating out of this structure into the nasal mesenchyme (Fig. 2 A, arrows) and along vomeronasal fibers of E12.5 embryos (Fig. 2 A, arrowheads). To establish whether Met-positive cells were GnRH-1-migrating neurons, double immunohistochemical stainings were performed. Double-labeling experiments for GnRH-1 and Met indicated that, at E12.5 (Fig. 2 B,C) and E14.5 (data not shown), the majority of GnRH-1 neurons were Met immunopositive (Fig. 2 B,C, arrows), as revealed by merged single confocal planes (Fig. 2 C, inset, arrows). Once within the brain, it was difficult to determine whether GnRH-1 neurons maintained Met expression, because of the high level of expression of this receptor in other CNS cells (data not shown). However, at postnatal day 10, Met expression was broadly downregulated throughout the brain, although there was evidence of discrete Met staining within the hypothalamus (Fig. 2 D, arrowheads). This corresponds to a stage when the GnRH-1 migratory process is over. Double labeling for GnRH-1 (Fig. 2 D–F, arrows) and Met (arrowheads) at PN10 revealed no coexpression between the two antigens, as shown by single confocal planes (Fig. 2 E,F). Thus, Met immunoreactivity is associated with migrating GnRH-1 neurons, being downregulated once these cells complete their migration.
HGF/Met expression in nasal explants
The HGF/Met expression pattern observed in nasal regions during development together with results from previous studies (Sonnenberg et al., 1993; Thewke and Seeds, 1996; Powell et al., 2001) suggest that HGF may have a role in regulating the GnRH-1 migratory process. In this context, nasal explants represent a valuable tool to separate spatial from temporal cues and focus on the properties of GnRH-1 neurons by controlling extracellular influences (Fueshko and Wray, 1994; Giacobini et al., 2004). It has been shown previously that the migrational pattern of GnRH-1 neurons observed in vivo reproducibly occurs in nasal explants in vitro; with a shift in location of the GnRH-1 cell population from the olfactory pit epithelia (OPEs) to the edge of the main tissue mass occurring from 1 to 3 div and continuing to more distant sites from 3 to 7 div (Fueshko and Wray, 1994).
To use nasal explants for functional studies, we first verified that this system retained expression of HGF and its receptor similar to the in vivo expression pattern. At 3 div, the majority of GnRH-1 neurons are located in the inner tissue mass of the explant, but some have started to migrate out into the periphery of the explant (Fig. 3 B, arrows). At this stage, HGF immunoreactivity was robustly expressed in the submucosa adjacent to the OPE, in the midline cartilage, and in mesenchymal cells located at the border between the inner tissue mass and the periphery, coinciding with the site at which GnRH-1 neurons and olfactory axons exit (Fig. 3 C, asterisks). This latter region corresponds to the frontonasal mesenchyme, also known as nasal/forebrain junction (n/fb J) in vivo. The expression pattern of Met receptor was then examined at the same in vitro stage. Met coexpressed with migrating GnRH-1 neurons as well as with the olfactory neurons in the OPE (Fig. 3 D, arrowhead). At 7 div, GnRH-1 neurons are located in the periphery of nasal explants in close association with the olfactory fibers. At this stage, the majority of GnRH-1 neurons expressed Met (Fig. 3 E,F, arrowheads), although some Met-positive/GnRH-1-negative cells were detected as well (Fig. 3 E, arrows). In addition, olfactory axons, along which GnRH-1 neurons migrated, exhibited Met staining (Fig. 3 E). Hence, consistent with in vivo results, Met receptor demarcated the olfactory system and the migrating GnRH-1 cell population.
tPA is expressed in migrating GnRH-1 neurons in vitro
tPA and uPA are serine proteases that, in addition to other proteases related to blood coagulation factor XII, have been shown to cleave and activate pro-HGF (Mars et al., 1993). Moreover, PAs expression is most pronounced during cell migration and axonal outgrowth processes in the developing nervous system (Seeds et al., 1997). Thus, tPA and uPA expression by GnRH-1 neurons was evaluated in vitro.
Single GnRH-1 cells were removed from nasal explants at 4.5 and 28 div (Fig. 4 A–C, arrows), two in vitro stages representative of GnRH-1 cells during migratory and postmigratory phases, respectively (Fueshko and Wray, 1994). cDNA pools were examined for tPA and uPA transcripts by single-cell RT-PCR. At 4.5 div, the majority of GnRH-1 cells (four of five) expressed tPA but not uPA transcripts. By 28 div, all GnRH-1 neurons (n = 5) were negative for both transcripts (Fig. 4 D). Double immunofluorescence for GnRH-1 and tPA was performed in nasal explants at 4.5 div (Fig. 4 E). These experiments revealed coexpression of the antigens (Fig. 4 E, arrows, inset) as well as expression of tPA along olfactory axons, confirming previous in situ hybridization studies (Thewke and Seeds, 1996). Thus, immunocytochemical experiments confirmed single-cell PCR results showing that tPA is expressed in GnRH-1 cells in a temporal window limited to the neuronal migratory process.
Nasal explants release bioactive HGF in the culture medium
The expression analyses demonstrate that HGF protein is present in nasal regions both in vivo and in vitro with a temporal and spatial pattern to impact GnRH-1/olfactory system development. The form of HGF observed in E12 noses by Western blots is indicative of a biologically active protein. To test whether the embryonic nasal region is able to release functional HGF, nasal explant CM was collected, and a typical scatter assay was performed, taking advantage of the Met-expressing MDCK cell line. In the absence of HGF, these cells grow in compact colonies (Fig. 5 A, inset, arrows). The addition of 10 ng/ml HGF for 48 h to MDCK cultures induced a typical change in morphology of MDCK cells and a scatter response resulting in cell dispersion (Fig. 5 B, inset, arrows). CM from 3 div nasal explants also enhanced migration capacity (Fig. 5 C), which was blocked by the addition of HGF-neutralizing antibody (5 μg/ml) (Fig. 5 D). The scatter response was quantified by measuring the number of MDCK cells contained within each counting frame (see Materials and Methods) (Fig. 5 E, arrows). This number decreases as a function of cell dispersion after increased migratory activity. Quantification of the scatter response showed more than a 50% reduction in the cell number contained within each counting frame in HGF- and CM-treated groups compared with control conditions (Fig. 5 E).
Anti-HGF disrupts GnRH-1 neuronal migration and olfactory axon outgrowth
To determine the role of endogenous HGF on GnRH-1/olfactory system development, the explants were treated with anti-HGF (5 μg/ml). The same concentration of this antibody has been used in previous studies to neutralize the activity of HGF (Powell et al., 2001; Giacobini et al., 2002). Nasal explants were treated from 3 to 6 div, a temporal window characterized by massive olfactory axonal growth and GnRH-1 neuronal migration from the inner tissue mass to the periphery of the explant (Fueshko and Wray, 1994). Nasal explants were fixed at 7 div and stained for GnRH-1 and peripherin, which stains the olfactory system (Fig. 6 A). No significant differences were found in total number of GnRH-1 cells inside the inner tissue mass (control, 64 ± 9; n = 18; anti-HGF treatment, 74 ± 9; n = 20) or in the periphery of the explant (control, 176 ± 22; n = 20; anti-HGF treatment, 150 ± 18; n = 21). No changes in GnRH-1 cell number after treatment suggests that mitogenic and survival effects of HGF on GnRH-1 neurons are unlikely. However, application of anti-HGF severely stunted the migration of GnRH-1 cells (Fig. 6 B–D, arrowheads, insets) and the peripherin-fiber network in the periphery of the explant (Fig. 6 C,D, arrows, insets). Quantitative assessment of olfactory fiber outgrowth revealed a significant reduction in the maximum distance from the border of the explant after treatment with anti-HGF (p < 0.001; control, 1188 ± 47 μm; n = 17; anti-HGF treatment, 996 ± 95 μm; n = 16). The complex nature of the fiber network prevented quantification of individual fiber lengths. Therefore, the observed reduced distance of the peripherin-positive fiber bundles could be the result of a defect in olfactory axon elongation or of an altered orientation of the olfactory axons.
For GnRH-1 cell migrational measurements, a caliber with a series of concentric arcs separated by a uniform distance (200 μm) was overlaid on the digitized image, and the number of cells in each zone was counted (Fig. 6 A). In addition to changes in olfactory axons, GnRH-1 neurons were closer to the border of the explant in the anti-HGF treated group compared with controls (Fig. 6 B–D) (p < 0.005). In control explants, 14% of the entire GnRH-1 cell population was dispersed beyond zone 4 (>600 μm from the border of the explant tissue mass), whereas when explants received anti-HGF, only 6% of GnRH-1 neurons migrated >600 μm into the periphery. In treated explants, GnRH-1 neurons also displayed an abnormal migratory behavior. In control conditions, the majority of migrating cells were uniformly oriented in a proximal-to-distal direction (Fig. 6 C, inset), whereas in the presence of HGF antibody, such polarized direction was lost, and GnRH-1 neurons orientation appeared more random (Fig. 6 D, inset). Although “randomly” oriented, the migrating neurons maintained contact with peripherin-positive fibers, which also appeared more entwined and less directionally oriented after anti-HGF treatment (Fig. 6 C,D). Thus, we were unable to determine whether the loss of orientation of GnRH-1 neurons was attributable to cell-autonomous mechanisms or instead directly dependent on altered outgrowth of olfactory axons.
Exogenous HGF increases GnRH-1 cell migration in nasal explants
We next evaluated the effect of exogenous HGF on GnRH-1 cell migration and/or on olfactory axon outgrowth. HGF treatment did not affect the total number of GnRH-1 neurons compared with controls (control, 221 ± 46; n = 11; HGF-treated, 204 ± 19; n = 10). However, a significant shift in the location of GnRH-1 neurons was noted in HGF-treated cultures (Fig. 7) (p < 0.001). In this group, 21% of GnRH-1 neurons in the periphery of the explant migrated beyond zone 5 (> 800 μm from the edge of the main tissue mass), compared with controls, which displayed only 8% of the GnRH-1 population in this same compartment. To determine whether HGF had an effect on olfactory axon outgrowth as well, the mean maximum network outgrowth of peripherin fibers was analyzed. Quantitative analysis revealed that the extent of fiber outgrowth was similar among HGF-treated and control explants (control, olfactory axon outgrowth: 1215 ± 57 μm; n = 11; HGF-treated, olfactory axon outgrowth: 1292 ± 48; n = 10), and directionality was maintained in all explant groups (data not shown).
HGF is a guidance signal for migrating GnRH-1 neurons
The HGF expression pattern suggests that GnRH-1 neurons may follow this diffusible molecule as they move from the VNO toward the nasal/forebrain junction. If this were the case, then addition of an exogenous source of HGF in a direction opposite to the normal migratory pathway (i.e., the rostral tip of the nose) should disrupt or delay GnRH-1 neuronal migration. To test this hypothesis, functional experiments were performed by coculturing for 24 h cell aggregates of HGF-transfected cells, together with parasagittal slices of whole heads of E12.5 mice. This embryonic age corresponds to a stage in which GnRH-1 neurons span from the VNO to the nasal/forebrain junction, on their way into the CNS (Wray, 2002).
MDCK EGFP–HGF stable transfectants were generated using the Tet-Off expression system (Gossen and Bujard, 1992). Clones were screened through EGFP imaging (Fig. 8 A, top left). When cells were cultured in the presence of 1 μg/ml Dox, EGFP and HGF expressions were turned off (Fig. 8 A, top right).
To verify HGF biosynthesis in these cells, total extracts were run under nonreducing conditions and immunoblotted with anti-VSVG (Fig. 8 B). Tet-Off MDCK cells expressing EGFP and HGF were grown in the absence (lane 1) or presence (lane 2) of Dox. Tet-Off MDCK cells expressing only EGFP were used as negative control (lane 3). Transfected HGF was identified as pro-HGF inside the cells and was expressed only in the absence of Dox. In addition, the ability to produce and release active HGF in the culture medium was tested by using the scatter assay. CM was collected from Tet-Off cells grown with or without Dox for several days, diluted 1:1 with fresh culture medium, and applied onto MDCK cells (Fig. 8 A, bottom).
HGF-releasing cell aggregates were placed at the tip of the nose (Fig. 9 C,D). The mean numbers of cells that were GnRH-1 immunoreactive per slice (± SEM) were 273 ± 21 (control, 0 div; n = 6), 301 ± 16 (control, 1 div, with Dox; n = 6), 285 ± 19 (HGF, 1 div, without Dox; n = 4), with no significant difference among groups (p > 0.05) (Fig. 9 B). The changing positions of GnRH-1-ir neurons from nasal compartment to the CNS between day 0 and day 1 control slices provided evidence of migration in vitro (Fig. 9 B). Slices grown for 1 div in the presence of EGFP–HGF cell aggregates silenced with Dox showed a broad distribution of GnRH-1 cells from the nasal compartment (Fig. 9 E, arrowheads) to the basal forebrain (Fig. 9 E, arrows). GnRH-1 neurons migrating through the nasal mesenchyme displayed a bipolar morphology and were visible as streams of neurons directed toward the cribriform plate (Fig. 9 E, inset). At this stage, 32% of the total GnRH-1 neurons entered the brain (Fig. 9 B,E, arrows). In contrast, GnRH-1 cells displayed an atypical migratory behavior when embryonic slices were cocultured with EGFP–HGF-expressing cell aggregates cultured in the absence of Dox (Fig. 9 D,F). The vast majority of GnRH-1 neurons (90%) did not reach the forebrain (Fig. 9 B,F, arrowheads), as a result of accumulation in the nasal compartment (Fig. 9 F, inset). Significant differences in GnRH-1 cell distribution between groups treated with and without Dox were observed both in the nose and in the CNS (p < 0.05).
Intense peripherin immunoreactivity was found on olfactory/vomeronasal axons that crossed the nasal mesenchyme and projected into the forebrain in cultures both with and without Dox (Fig. 9 G,H). The olfactory fiber network in the nose as well as in the forebrain did not appear to be disrupted when HGF-releasing cell aggregates were cocultured with embryonic slice cultures (Fig. 9 G,H).
Fewer GnRH-1 neurons are present in the tPA/uPA deficient mice
Early embryonic lethality of met mutants has prevented in vivo studies on these mice to determine the functional role of HGF in GnRH-1 neuron development (Bladt et al., 1995; Schmidt et al., 1995; Uehara et al., 1995). Therefore, we investigated the effect of deletion of tPA and uPA genes on the number of GnRH-1 neurons in adult brains. These serine proteases have been shown to activate the progrowth factor HGF (Mars et al., 1993). Moreover, previous studies showed that mice with combined deficiencies of tPA and uPA are subfertile and display reduced gonadotropin-induced ovulation efficiency (Carmeliet et al., 1994; Leonardsson et al., 1995). The total number of GnRH-1 neurons was compared in 60- to 90-d-old WT male mice (n = 3) and age and sex-matched double-KO mice for tPA and uPA genes (n = 3). Analysis revealed a significant reduction in GnRH-1 cell number in the brains of mutants compared with WT (WT, 642 ± 16; tPA−/−:uPA−/−, 422 ± 24; p < 0.001). A reduction of ∼35% was found in the KO animals. Figure 10, A and B, shows representative stainings for GnRH-1 neurons (arrows, single GnRH-1 cells; arrowhead, cluster of GnRH-1 cells) in the diagonal band of Broca (dbb) of the hypothalamus of WT and double-KO mice. At this level, numerous GnRH-1 neurons are normally detected (Fig. 10 A). In contrast, few GnRH-1 neurons are detected at this level in KO mice (Fig. 10 B). The median eminence of tPA−/−:uPA−/− brains was also sparsely innervated by GnRH-1-immunoreactive terminals (Fig. 10 D, arrowhead, inset) compared with WT (Fig. 10 C, arrowhead, inset).
Development of the olfactory and GnRH-1 neuroendocrine systems is intimately entwined in early embryogenesis (Wray, 2002). The mechanisms directing the initiation of cell migration and olfactory axon extension from nose to forebrain are unclear but likely require specific motogenic and guidance cues. In this report, we show that HGF and Met are expressed in a spatiotemporal pattern to impact GnRH-1/olfactory system development. Functional analysis supports the notion that HGF plays an important role in regulating GnRH-1 neuronal migration across nasal regions toward the CNS during embryogenesis.
Previous studies have shown that c-met and tPA mRNAs are expressed in olfactory epithelium, whereas HGF transcript localized to the surrounding nasal mesenchyme starting at E11 in mouse (Sonnenberg et al., 1993; Thewke and Seeds, 1996), when the GnRH-1/olfactory systems are in their initial stages of development. Here, biologically active HGF and Met protein expression was documented in nasal regions as early as E12. The majority of GnRH-1 neurons located in the nasal compartment were found to be Met immunopositive with expression correlated with migration. Using RT-PCR, c-met mRNA was detected in laser-captured tissues of OE and VNO at E14.5, confirming previous in situ hybridization studies (Sonnenberg et al., 1993; Thewke and Seeds, 1996). Moreover, immunohistochemistry coupled with confocal microscopy revealed that Met protein is expressed along NCAM-positive olfactory fibers during embryonic development. Thus, the spatiotemporal expression of HGF and Met receptor in nasal regions correlates with migration of GnRH-1 neurons toward the CNS and development of the olfactory sensory system.
To determine the functional role of HGF in the development of the GnRH-1 system, we took advantage of an in vitro model, nasal explants, which has been successfully used for other functional studies (Fueshko et al., 1998; Kramer and Wray, 2000; Giacobini et al., 2004). These explants maintain large numbers of GnRH-1 neurons, migrating in a manner similar to that observed in vivo, as well as directed olfactory axon outgrowth (Fueshko and Wray, 1994). Expression of Met in the olfactory system and in primary GnRH-1 neurons in nasal explants was similar to that observed in vivo. HGF was expressed as early as 3 div, a stage of active cell migration and olfactory axon outgrowth, and the expression paralleled HGF transcript distribution described previously in vivo (Sonnenberg et al., 1993; Thewke and Seeds, 1996). HGF immunoreactivity was observed in the olfactory mucosa surrounding the OE, in the nasal midline cartilage, and in the frontonasal mesenchyme, which is the region apposed to the ventromedial forebrain before dissection. Olfactory pathway development depends on induction between the frontonasal mesenchyme and adjacent olfactory epithelia (LaMantia et al., 2000). Interestingly, HGF mRNA has been shown to be unevenly distributed in the nasal mesenchyme during embryonic development, being expressed in a gradient with higher levels toward the forebrain (Sonnenberg et al., 1993). A similar expression pattern has been shown for SDF-1 in these regions (Schwarting et al., 2006). This transcript is expressed in a steep gradient in the developing nasal mesenchyme, being lower proximal to the VNO and greater at the nasal/forebrain junction. SDF-1 was shown to be important for the migration of GnRH-1 neurons (Toba et al., 2004; Schwarting et al., 2006). Other studies demonstrated that SDF-1 and HGF regulate recruitment of mesenchymal stem cells toward damaged tissues (Ji et al., 2004; Urbanek et al., 2005) and that HGF chemotactic response could be potentiated by SDF-1 (Son et al., 2006). Similarly, muscle progenitor cells migrate toward SDF-1-expressing targets with CXCR4 (CXC receptor 4), the SDF-1 receptor, and Gab1 [Grb2 (growth factor receptor-bound protein 2)-associated binding protein 1], the adaptor molecule that transmits Met signaling, cooperating to control this process (Vasyutina et al., 2005). Thus, we cannot rule out cross talk between the HGF and SDF-1 signaling pathways in regulating directed migration of GnRH-1 neurons from the VNO to the brain.
In addition to Met, primary GnRH-1 neurons express tPA. Interestingly, this expression pattern correlated with migration of these cells, being downregulated in postmigratory GnRH-1 neurons. Previous studies showed that tPA transcript is expressed in migrating neurons crossing the nasal mesenchyme during early stages of embryonic development (Friedman and Seeds, 1994). Here, we suggest that at least part of these cells are GnRH-1 neurons and that these cells display the molecular machinery to activate HGF in the immediate vicinity of its c-met receptor, thus initiating a cell signaling cascade that influences cell movement.
Our in vitro experiments demonstrated that biologically active HGF was released into the medium of nasal explants. In the motogenic assay, this medium induced a scatter response in MDCK cells, which was blocked by HGF-neutralizing antibody. HGF was released into the nasal explant medium in the initial period of GnRH-1 neuronal migration and olfactory axon elongation. Anti-HGF treatment significantly stunted GnRH-1 migratory behavior and olfactory axon outgrowth (elongation or orientation), supporting endogenous HGF acting as a motogen on GnRH-1 neurons and a growth promoter for olfactory axons. When GnRH-1 cells were subject to exogenous uniform HGF concentrations, the cells did migrate farther and kept their spatial orientation. No effect was observed in terms of olfactory axon growth, in contrast with olfactory axon changes detected in the blocking-function experiments. This apparent discrepancy may result because olfactory axon extension cannot be stimulated above an intrinsic limit or a limit imposed by general fibroblast outgrowth. Therefore, exogenous HGF might be insufficient to promote additional elongation of olfactory axons.
After HGF exposure, parallel changes were not detected in the GnRH-1 and olfactory system. Thus, the changes observed in GnRH-1 cell migration do not appear to be an effect dependent on alterations in olfactory axon outgrowth and suggest that the HGF effect on the motility of GnRH-1 neurons is in fact cell autonomous. It is important to note that, although anti-HGF treatment induced a striking accumulation of cells and olfactory fibers closer to the nasal explant tissue mass, it did not prevent GnRH-1 cells and peripherin-positive fibers from moving/extending into the periphery of the explant. Likely, residual HGF and perhaps other factors present in the nasal explants contribute to the initiation of cell movement and olfactory axon outgrowth.
We showed previously that HGF acts as a chemoattractant for immortalized GnRH-1 cells (Giacobini et al., 2002). Organotypic slices have been shown previously to contain the conditions necessary to allow GnRH-1 migration from the developing olfactory system to the basal forebrain while keeping intact their migratory route (Tobet et al., 1996). Thus, retained in these slices are the guidance cues for GnRH-1/olfactory system development that are clearly lacking in immortalized cells. In addition, some phenotypic/behavioral traits of the cell lines could be a consequence of the immortalization procedure which may alter the expression pattern and the activity of such cells (Martinez de la Escalera and Clapp, 2001). Thus, to evaluate whether HGF acts as a guidance cue during the migration of primary GnRH-1 neurons, we used tissue slices prepared from mouse embryos (Tobet et al., 1996; Bless et al., 2000). Tissue slices that maintain connection between forebrain and nasal compartment were generated from E12.5 whole heads. Embryonic slices were cocultured with aggregates of EGFP–HGF-transfected cells. In these experiments, after 1 d in vitro, GnRH-1 neurons accumulated in the nasal region when slices were cocultured with HGF-releasing transfected cell aggregates placed in an opposite direction with respect to the normal GnRH-1 migratory pathway.
The disruption of GnRH-1 cell motility observed in these experiments may result from their exposure to two attractive HGF gradients (one endogenous and one exogenous) acting from two opposite sites or from threshold concentrations of exogenous HGF, which fail to generate a motogenic response. In these in vitro coculture experiments, no alterations were observed in peripherin-positive olfactory/vomeronasal nerve axons among treatment groups, suggesting that the observed HGF-induced GnRH-1 neuronal migration defect is cell autonomous. However, it should also be noted that, at E12.5, olfactory fibers could no longer be sensitive to HGF stimulation, as they have started to contact their tissue target (i.e., the presumptive olfactory bulbs). In fact, it is known that at E11 in the mouse, axons begin to extend from the OE, forming small fascicles which pierce the basal lamina (Marin-Padilla and Amieva, 1989).
HGF-induced activity requires proteolytic processing, which is, at least in part, operated through enzymatic cleavage by PAs (Naldini et al., 1992; Mars et al., 1993). Mice carrying combined deficiencies for tPA and uPA genes are subfertile (Carmeliet et al., 1994), consistent with the hypothesis that the lack of tPA and uPA genes prevents active HGF from promoting GnRH-1 neuronal migration into the forebrain. To examine this possibility, the number of GnRH-1 neurons in adult brains of tPA−/−:uPA−/− mice was compared with those of WT animals. The lack of tPA and uPA clearly affected the size of the GnRH-1 neuronal population, decreasing the number of GnRH-1 cells detected in brain. Whether this effect is directly related to loss of active HGF on GnRH-1 neurons or to additional mechanisms affecting GnRH-1 differentiation or cell survival will require additional investigations.
Very little is known about early interactions between migrating GnRH-1 neurons and the nasal mesenchyme, which lead to proper initial movement from the presumptive VNO toward the rostral forebrain. Our results are consistent with a role for HGF as a motogen as well as a chemotactic signal for developing GnRH-1 neurons, acting to positively modulate cell–cell and cell–ECM interactions during the early aspects of their migratory process.
This work was supported by Compagnia di San Paolo (Neurotransplant Project 2004.2019), Ricerca Scientifica Applicata Comitato Interministeriale Programmazione Economica A23 Regione Piemonte, and Fondo per gli Investimenti della Ricerca di Base Grant RBNE01WY7P (Italy). We thank Andree Reuss for her help in generating nasal explants.
- Correspondence should be addressed to Dr. Paolo Giacobini, Department of Human and Animal Biology, Via Accademia Albertina 13, 10123 Torino, Italy.