[Association between fibroblast growth factor 3 polymorphism and non-syndromic oral clefting]

OBJECTIVE

To investigate the association between fibroblast growth factor 3 (FGF3) gene rs4980700 and rs4631909 polymorphism and non-syndromic oral clefting (NSOC).

METHODS

Blood samples from 186 NSOC patients, patients' parents and 200 controls were collected. DNA was extracted and PCR-restriction fragment length polymorphism (PCR-RFLP) was used to identify genotypes of the samples. Case-control analyses and transmission disequilibrium test (TDT) and family based association test (FBAT) analyses were also carried out.

RESULTS

In case-control analysis, there were significant differences in rs4980700 genotype and allele among NSOC patients compared with the control group (P < 0.05) and there were significant differences in rs4631909 genotype and allele among NSOC patients compared with the control group (P < 0.05), but no difference in cleft palate only (P = 0.49). In TDT, the G allele of rs4980700 had an overtransmission (P < 0.05) and the C allele of rs4631909 had an overtransmission (P < 0.05) in NSOC. FBAT analysis also showed a significant association between FGF3 gene rs4980700, rs4631909 polymorphism and NSOC.

CONCLUSION

FGF3 gene rs4980700 and rs4631909 polymorphism were associated with NSOC.

FGF signaling regulates otic placode induction and refinement by controlling both ectodermal target genes and hindbrain Wnt8a

The inner ear epithelium, with its complex array of sensory, non-sensory, and neuronal cell types necessary for hearing and balance, is derived from a thickened patch of head ectoderm called the otic placode. Mouse embryos lacking both Fgf3 and Fgf10 fail to initiate inner ear development because appropriate patterns of gene expression fail to be specified within the pre-otic field. To understand the transcriptional "blueprint" initiating inner ear development, we used microarray analysis to identify prospective placode genes that were differentially expressed in control and Fgf3(-)(/)(-);Fgf10(-)(/)(-) embryos. Several genes in the down-regulated class, including Hmx3, Hmx2, Foxg1, Sox9, Has2, and Slc26a9 were validated by in situ hybridization. We also assayed candidate target genes suggested by other studies of otic induction. Two placode markers, Fgf4 and Foxi3, were down-regulated in Fgf3(-)(/)(-);Fgf10(-)(/)(-) embryos, whereas Foxi2, a cranial epidermis marker, was expanded in double mutants, similar to its behavior when WNT responses are blocked in the otic placode. Assays of hindbrain Wnt genes revealed that only Wnt8a was reduced or absent in FGF-deficient embryos, and that even some Fgf3(-)(/)(-);Fgf10(-)(/+) and Fgf3(-)(/)(-) embryos failed to express Wnt8a, suggesting a key role for Fgf3, and a secondary role for Fgf10, in Wnt8a expression. Chick explant assays showed that FGF3 or FGF4, but not FGF10, were sufficient to induce Wnt8a. Collectively, our results suggest that Wnt8a provides the link between FGF-induced formation of the pre-otic field and restriction of the otic placode to ectoderm adjacent to the hindbrain.

INT-2 gene amplification in differentiated human thyroid cancer

Oncogene amplification is frequent in many epithelial tumors and often associated with advanced tumor progression. In different epithelial neoplasias it helps to provide prognostic information on individual patients. The present study was performed to evaluate the hitherto unknown prevalence of INT-2 gene amplification and its potential usefulness as prognostic marker in patients with human thyroid cancer. We used differential quantitative polymerase chain reaction and fluorescent DNA technique as a reliable method to detect low copy-number amplification of oncogenes from archival carcinoma specimens. Sequences from the int-2 gene and the single copy gamma-interferon gene were amplified simultaneously by PCR and quantified on a fluorescence activated sequencer. Native tumor tissue from 63 patients with differentiated thyroid cancer (43 papillary, 3 oncocytary, and 17 follicular) and from 12 goiters was analyzed by differential quantitative polymerase chain reaction. The study group contained many far advanced tumors. 40% of tumors were recurrent, 35% were staged T4 tumors and 70% presented with lymph node metastases. The prevalence of INT-2 amplification was 12% for follicular and 7% for papillary carcinomas. In goiter tissue no amplification was found. Amplification was only 2-4fold in positive cases. Low grade amplification is of no apparent importance in differentiated thyroid cancer.

Endoderm-derived Fgf3 is necessary and sufficient for inducing neurogenesis in the epibranchial placodes in zebrafish

In vertebrates, epibranchial placodes are transient ectodermal thickenings that contribute sensory neurons to the epibranchial ganglia. These ganglia innervate internal organs and transmit information on heart rate, blood pressure and visceral distension from the periphery to the central nervous system. Despite their importance, the molecular mechanisms that govern the induction and neurogenesis of the epibranchial placodes are only now being elucidated. In this study, we demonstrate that endoderm is required for neurogenesis of the zebrafish epibranchial placodes. Mosaic analyses confirm that endoderm is the source of the neurogenic signal. Using a morpholino knockdown approach, we find that fgf3 is required for the majority of placode cells to undergo neurogenesis. Tissue transplants demonstrate that fgf3 activity is specifically required in the endodermal pouches. Furthermore, ectopic fgf3 expression is sufficient for inducing phox2a-positive neurons in wild-type and endoderm-deficient embryos. Surprisingly, ectodermal foxi1 expression, a marker for the epibranchial placode precursors, is present in both endoderm-deficient embryos and fgf3 morphants, indicating that neither endoderm nor Fgf3 is required for initial placode induction. Based on these findings, we propose a model for epibranchial placode development in which Fgf3 is a major endodermal determinant required for epibranchial placode neurogenesis.

Comparative genomics on mammalian Fgf3-Fgf4 locus

The CCND1-ORAOV1-FGF19-FGF4-FGF3-TMEM16A-FADD-PPFIA1-CTTN (EMS1) locus at human chromosome 11q13.3 is amplified in head and neck tumors, esophageal cancer, Kaposi's sarcoma, bladder tumors, breast cancer, and liver cancer. Fgf4 mRNA is expressed in embryonic stem (ES) cells depending on Sox2 and Pou5f1 (Oct3/Oct4) transcription factors, and in myotomes and limb bud AER depending on MyoD (or Myf5) and GATA transcription factors. Here, rat Fgf3 and Fgf4 complete coding sequences were determined by using bioinformatics. Multiple errors, including one-base insertion and 22-base deletion, were identified within the coding region of rat Fgf4 RefSeq (NM_053809.1 or AB079673.1). Rat Fgf3 and Fgf4 genes, consisting of three exons, were clustered in tail-to-head manner with an interval of about 16 kb. CUTL1 (CCAAT-displacement protein, CDP) and NKX2-5 binding sites and TATA box within 5'-flanking promoter region were conserved among human, rat and mouse Fgf3 orthologs. MYOD and MYOG (Myogenin) binding sites and TATA box within 5'-flanking promoter region as well as GATA, MYOD, SOX2 and POU5F1 binding sites within exon 3 were conserved among mammalian Fgf4 orthologs. Human FGF3 and FGF4 genes were clustered in tail-to-head manner with an interval of about 35 kb. Major repetitive sequence (FGF34Rep1) and minor repetitive sequence (FGF34Rep2) were identified within human FGF3-FGF4 gene cluster. FGF34Rep1 were clustered within the FGF3-FGF4 locus as well as around the IL28RA locus (1p36.11) and the NFAM1 locus (22q13.2). FGF34Rep2 was characterized by the CCA(T/C) repeats. This is the first report on comparative genomics analyses on the Fgf3-Fgf4 locus within human, rat and mouse genomes.

Zebrafish gcmb is required for pharyngeal cartilage formation

The glial cells missing (gcm) gene in Drosophila encodes a GCM-motif transcription factor that functions as a binary switch to select between glial and neuronal cell fates. To understand the function of gcm in vertebrates, we isolated the zebrafish gcmb and analyzed the function of this gene using antisense morpholino oligonucleotides against gcmb mRNA (gcmb-MO) and transgenic overexpression. Zebrafish gcmb is expressed in the pharyngeal arch epithelium and in cells of the macrophage lineage. gcmb-MO-injected larvae show significantly reduced branchial arch cartilages. fgf3-MO-injected larvae display a similar phenotype to that of gcmb-MO-injected larvae with respect to the lack of pharyngeal cartilage formation. In addition, gcmb expression in the pharyngeal arches is down-regulated in fgf3-MO-injected larvae. The gcmb transgenic larvae show a protrusion of the lower jaw and abnormal spatial arrangement of the pharyngeal cartilage elements. These results suggest that gcmb is required for normal pharyngeal cartilage formation in zebrafish and that its expression is dependent on fgf3 activity.

Regulation of neuronal death and calcitonin gene-related peptide by fibroblast growth factor-2 and FGFR3 after peripheral nerve injury: Evidence from mouse mutants

The presence and regulation of basic fibroblast growth factor and its high-affinity tyrosine kinase receptor FGFR3 in sensory neurons during development and after peripheral nerve injury suggest a physiological role of the fibroblast growth factor-2 system for survival and maintenance of sensory neurons. Here we investigated L5 spinal ganglia of intact and lesioned fibroblast growth factor-2 knock-out and FGFR3 knock-out mice. Quantification of sensory neurons in intact L5 spinal ganglia revealed no differences between wild-types and mutant mice. After sciatic nerve axotomy, the normally occurring neuron loss in wild-type mice was significantly reduced in both knock-out strains suggesting that fibroblast growth factor-2 is involved in neuronal death mediated via FGFR3. In addition, the number of chromatolytic and eccentric cells was significantly increased in fibroblast growth factor-2 knock-out mice indicating a transient protection of injured spinal ganglia neurons in the absence of fibroblast growth factor-2. The expression of the neuropeptide calcitonin gene-related peptide in sensory neurons of intact fibroblast growth factor-2 knock-out and FGFR3 knock-out mice was not changed in comparison to adequate wild-types. Fibroblast growth factor-2 wild-type and FGFR3 wild-type mice showed a lesion-induced decrease of calcitonin gene-related peptide-positive neurons in ipsilateral L5 spinal ganglia whereas the loss of calcitonin gene-related peptide-immunoreactive sensory neurons is reduced in the absence of fibroblast growth factor-2 or FGFR3, respectively. In addition, FGFR3 wild-type and knock-out mice displayed a contralateral reduction of the neuropeptide after axotomy. These results suggest that endogenous fibroblast growth factor-2 and FGFR3 are crucially involved in the regulation of survival and calcitonin gene-related peptide expression of lumbar sensory neurons after lesion, but not during development.

Pax8 and Pax2a function synergistically in otic specification, downstream of the Foxi1 and Dlx3b transcription factors

The vertebrate inner ear arises from an ectodermal thickening, the otic placode, that forms adjacent to the presumptive hindbrain. Previous studies have suggested that competent ectodermal cells respond to Fgf signals from adjacent tissues and express two highly related paired box transcription factors Pax2a and Pax8 in the developing placode. We show that compromising the functions of both Pax2a and Pax8 together blocks zebrafish ear development, leaving only a few residual otic cells. This suggests that Pax2a and Pax8 are the main effectors downstream of Fgf signals. Our results further provide evidence that pax8 expression and pax2a expression are regulated by two independent factors, Foxi1 and Dlx3b, respectively. Combined loss of both factors eliminates all indications of otic specification. We suggest that the Foxi1-Pax8 pathway provides an early 'jumpstart' of otic specification that is maintained by the Dlx3b-Pax2a pathway.

Zebrafish pax8 is required for otic placode induction and plays a redundant role with Pax2 genes in the maintenance of the otic placode

Vertebrate Pax2 and Pax8 proteins are closely related transcription factors hypothesized to regulate early aspects of inner ear development. In zebrafish and mouse, Pax8 expression is the earliest known marker of otic induction, and Pax2 homologs are expressed at slightly later stages of placodal development. Analysis of compound mutants has not been reported. To facilitate analysis of zebrafish pax8, we completed sequencing of the entire gene, including the 5' and 3' UTRs. pax8 transcripts undergo complex alternative splicing to generate at least ten distinct isoforms. Two different subclasses of pax8 splice isoforms encode different translation initiation sites. Antisense morpholinos (MOs) were designed to block translation from both start sites, and four additional MOs were designed to target different exon-intron boundaries to block splicing. Injection of MOs, individually and in various combinations, generated similar phenotypes. Otic induction was impaired, and otic vesicles were small. Regional ear markers were expressed correctly, but hair cell production was significantly reduced. This phenotype was strongly enhanced by simultaneously disrupting either of the co-inducers fgf3 or fgf8, or another early regulator, dlx3b, which is thought to act in a parallel pathway. In contrast, the phenotype caused by disrupting foxi1, which is required for pax8 expression, was not enhanced by simultaneously disrupting pax8. Disrupting pax8, pax2a and pax2b did not further impair otic induction relative to loss of pax8 alone. However, the amount of otic tissue gradually decreased in pax8-pax2a-pax2b-deficient embryos such that no otic tissue was detectable by 24 hours post-fertilization. Loss of otic tissue did not correlate with increased cell death, suggesting that otic cells dedifferentiate or redifferentiate as other cell type(s). These data show that pax8 is initially required for normal otic induction, and subsequently pax8, pax2a and pax2b act redundantly to maintain otic fate.

Chromatin-dependent E1A activity modulates NF-kappaB RelA-mediated repression of glucocorticoid receptor-dependent transcription

The role of chromatin-dependent regulatory mechanisms in the repression of glucocorticoid-dependent transcription from the murine mammary tumor virus (MMTV) promoter by p65 and E1A was investigated by using chromatin and transiently transfected reporters. The p65 RelA subunit of NF-kappaB represses MMTV expression on either transient or integrated reporters. In contrast, the viral oncoprotein E1A represses a transient but not an integrated MMTV. E1A repression is attenuated by chromatin, suggesting p65 but not E1A manipulates chromatin appropriately to inhibit the GR. Coexpression of p65 and E1A additively represses the transient MMTV but restores the transcriptional activation of the chromatin MMTV in response to glucocorticoids. This indicates that E1A has a dominant chromatin-dependent activity that attenuates repression by p65. E1A, p65, and GR bind the MMTV promoter, and chromatin remodeling enhances binding on both repressed and activated promoters. In addition, p65 requires Brg for repression of the integrated MMTV. This suggests that neither p65 repression nor E1A attenuation of repression results from an inhibition of remodeling that prevents transcription factor binding. Furthermore, p300/CBP is also required for both repression and attenuation by p65 and E1A. E1A and p65 mutants that do not bind p300/CBP are inactive, indicative of a requirement for p300/CBP-dependent complex formation for both repression and attenuation with chromatin. These data suggest that both the p65-dependent repression and the E1A-mediated attenuation of repression require the Brg1-dependent chromatin remodeling function and p300/CBP-dependent complex formation at a promoter assembled within chromatin.

An essential role for Fgfs in endodermal pouch formation influences later craniofacial skeletal patterning

Fibroblast growth factor (Fgf) proteins are important regulators of pharyngeal arch development. Analyses of Fgf8 function in chick and mouse and Fgf3 function in zebrafish have demonstrated a role for Fgfs in the differentiation and survival of postmigratory neural crest cells (NCC) that give rise to the pharyngeal skeleton. Here we describe, in zebrafish, an earlier, essential function for Fgf8 and Fgf3 in regulating the segmentation of the pharyngeal endoderm into pouches. Using time-lapse microscopy, we show that pharyngeal pouches form by the directed lateral migration of discrete clusters of endodermal cells. In animals doubly reduced for Fgf8 and Fgf3, the migration of pharyngeal endodermal cells is disorganized and pouches fail to form. Transplantation and pharmacological experiments show that Fgf8 and Fgf3 are required in the neural keel and cranial mesoderm during early somite stages to promote first pouch formation. In addition, we show that animals doubly reduced for Fgf8 and Fgf3 have severe reductions in hyoid cartilages and the more posterior branchial cartilages. By examining early pouch and later cartilage phenotypes in individual animals hypomorphic for Fgf function,we find that alterations in pouch structure correlate with later cartilage defects. We present a model in which Fgf signaling in the mesoderm and segmented hindbrain organizes the segmentation of the pharyngeal endoderm into pouches. Moreover, we argue that the Fgf-dependent morphogenesis of the pharyngeal endoderm into pouches is critical for the later patterning of pharyngeal cartilages.

vhnf1 integrates global RA patterning and local FGF signals to direct posterior hindbrain development in zebrafish

The vertebrate hindbrain is transiently divided along the anterior-posterior axis into seven morphologically and molecularly distinct segments, or rhombomeres, that correspond to Hox expression domains. The establishment of a proper 'hox code' is required for the development of unique rhombomere identities, including specification of neuronal fates. valentino (val), the zebrafish ortholog of mafB/Kreisler (Kr), encodes a bZip transcription factor that is required cell autonomously for the development of rhombomere (r) 5 and r6 and for activation of Hox group 3 gene expression. Recent work has demonstrated that the expression of val itself depends on three factors: retinoic acid (RA) signals from the paraxial mesoderm; fibroblast growth factor (Fgf) signals from r4; and variant hepatocyte nuclear factor 1 (vhnf1, also known as tcf2), a homeodomain transcription factor expressed posterior to the r4-5 boundary. We have investigated the interactions between these inputs onto val expression in the developing zebrafish hindbrain. We show that RA induces val expression via activation of vhnf1 expression in the hindbrain. Fgf signals from r4, acting through the MapK pathway, then cooperate with Vhnf1 to activate val expression and subsequent r5 and r6 development. Additionally, vhnf1 and val function as part of a multistep process required for the repression of r4 identity in the posterior hindbrain. vhnf1 acts largely independently of val to repress the r4 'hox code' posterior to the r4-5 boundary and therefore to block acquisition of r4-specific neuronal fates in the posterior hindbrain. However, vhnf1 is not able to repress all aspects of r4 identity equivalently. val is required downstream of vhnf1 to repress r4-like cell-surface properties, as determined by an 'Eph-ephrin code', by repressing ephrin-B2a expression in r5 and r6. The different requirements for vhnf1 and val to repress hoxb1a and ephrin-B2a, respectively, demonstrate that not all aspects of an individual rhombomere's identity are regulated coordinately.

Genome-wide profiling of oral squamous cell carcinoma

Oral squamous cell carcinoma (OSCC) is a common malignancy, the incidence of which is particularly high in some Asian countries due to the geographically linked areca quid (AQ) chewing habit. In this study, array-based comparative genomic hybridization was used to screen microdissected OSCCs for genome-wide alterations. The highest frequencies of gene gain were detected for TP63, Serpine1, FGF4/FGF3, c-Myc and DMD. The highest frequencies of deletion were detected for Caspase8 and MTAP. Gained genes, classified by hierarchical clustering, were mainly on 17q21-tel; 20q; 11q13; 3q27-29 and the X chromosome. Among these, gains of EGFR at 7p, FGF4/FGF3, CCND1 and EMS1 at 11q13, and AIB1 at 20q were significantly associated with lymph node metastasis. The genomic profiles of FHIT and EXT1 in AQ-associated and non-AQ-associated OSCCs exhibited the most prominent differences. RT-PCR confirmed the significant increase of TP63 and Serpine1 mRNA expression in OSCC relative to non-malignant matched tissue. A significant increase in Serpine1 immunoreactivity was observed from non-malignant matched tissue to OSCC. However, there was no correlation between the frequent genomic loss of Caspase 8 and a significant decrease in Caspase8 expression. These data demonstrate that genomic profiling can be useful in analysing pathogenetic events involved in the genesis or progression of OSCC.

Early requirement for fgf8 function during hindbrain pattern formation in zebrafish

Fibroblast growth factor (FGF) signaling is required for normal development of the vertebrate brain, including the isthmus and caudal regions of the hindbrain. Recent work in zebrafish has identified a requirement for the combination of fgf3 and fgf8 functions in specification of rhombomeres 5 and 6 (r5, r6), when evaluated at mid- and late somitogenesis stages. However, when examined earlier in development, during early somitogenesis stages, FGF8 alone is required to initiate r5 and r6 development. Both a mutation in fgf8 and injection of fgf8-targeted antisense morpholino-modified oligonucleotides result in suppression of genes normally expressed in r5 and r6 by the one- to two-somite stage. This expression recovers by the six-somite stage, and we propose that this recovery is a response to activation of fgf3 and to delayed accumulation of fgf8. These data demonstrate an early, nonredundant requirement for fgf8 function in hindbrain patterning.

Regulation of cranial suture morphogenesis

The cranial sutures are the primary sites of bone formation during skull growth. Morphogenesis and phenotypic maintenance of the cranial sutures are regulated by tissue interactions, especially those with the underlying dura mater. Removal of the dura mater in fetuses causes abnormal suture development and premature suture obliteration. The dura mater interacts with overlying tissues of the cranial vault by providing: (1) intercellular signals, (2) mechanical signals and (3) cells, which undergo transformation and migrate to the suture. The intercellular signaling governing suture development employs the fibroblast growth factors (FGFs). In rats during formation of the sutures in the fetus, FGF-1 is localized mainly in the dura mater, while other FGFs are expressed in the overlying tissues. By birth, FGF-2 largely replaces FGF-1 in the dura mater. FGFs present in the calvaria bind either the IIIb or IIIc mRNA splice variants of the FGF receptors (FGFRs) 1, 2, or 3. Monoclonal antibodies to the b variant of FGFR2 were used to determine the distribution of FGFR2IIIb during suture development and its extracellular localization. FGFR2IIIb is present in association with mature osteoblasts and osteogenic precursor cells of the suture in the fetus. Ectodomains of FGFR2IIIb, the products of proteolytic cleavage of the receptors, were present throughout the extracellular matrix of sutures resisting obliteration (coronal and sagittal), but absent from the core of sutures undergoing normal fusion (posterior intrafrontal). This observation is consistent with a possible mechanism, in which truncated receptors bind FGFs, thus regulating free FGF available to nearby cells. Mechanical signaling in the calvaria results from tensional forces in the dura mater generated during rapid expansion of the neurocranium. Posterior intrafrontal sutures of rats, which fuse between days 16 and 24, were subjected to cyclical tensional forces in vitro. Significant delay in the timing of suture fusion and increases in the expression domains of FGFR1 and 2 were observed, demonstrating the sensitivity of suture patency to mechanical signals and a possible role of the FGF system in mediating such stimuli. Finally, cells of the dura mater beneath the intrafrontal and sagittal sutures were observed to undergo a morphological transformation to a dendritic morphology and migrate into the suture mesenchyme between days 10 and 16 of development. This process may participate in suture and bone morphogenesis and influence the patency of the sutures along the anterior-posterior axis.

Progression of pregnancy-dependent mouse mammary tumors after long dormancy periods. Involvement of Wnt pathway activation

Mouse mammary tumor virus (LA) induces pregnancy-dependent mammary tumors that progress toward autonomy. Here we show that in virgin females, pregnancy-dependent tumor transplants are able to remain dormant for up to 300 days. During that period, these tumors synthesize DNA, express high levels of estrogen and progesterone receptors (ER+PR+) and are able to resume growth after hormone stimulation. Surprisingly, in a subsequent transplant generation, all these tumors are fully able to grow in virgin females, they express low levels of ER and PR (ER-PR-) and have a monoclonal origin; i.e., show all of the features we have described previously in pregnancy-independent tumors. Histologically, mouse mammary tumor virus (LA)-induced tumors are morphologically similar to genetically engineered mouse (GEM) mammary tumors that overexpress genes belonging to the Wnt pathway. Interestingly, in the virus-induced neoplasias, pregnancy-independent passages arising after a dormant phase usually display a lower level of glandular differentiation together with epithelial cell trans-differentiation, a specific feature associated to Wnt pathway activation. In addition, dormancy can lead to the specific selection of Int2/Fgf3 mutated and overexpressing cells. Therefore, our results indicate that during hormone-dependent tumor dormancy, relevant changes in cell population occur, allowing rapid progression after changes in the animal internal milieu.

fgf17b, a novel member of Fgf family, helps patterning zebrafish embryos

Fibroblast growth factors (Fgfs) play important roles in the pattern formation of early vertebrate embryos. We have identified a zebrafish ortholog of human FGF17, named fgf17b. The first phase of fgf17b expression occurs in the blastodermal margin of late blastulae and in the embryonic shield of early gastrulae. The second phase starts after the onset of segmentation, mainly in the presomitic mesoderm and newly formed somites. Injection of fgf17b mRNA into one-cell embryos induces expression of the mesodermal marker no tail (ntl) and rescues ntl expression suppressed by overexpression of lefty1 (lft1). Overexpression of fgf17b dorsalizes zebrafish gastrulae by enhancing expression of chordin (chd), which is an antagonist of the ventralizing signals BMPs. In addition, overexpression of fgf17b posteriorizes the neuroectoderm. Simultaneous knockdown of fgf17b and fgf8 with antisense morpholinos results in reduction of chd and ntl. Knockdown of fgf17b can alleviate inhibitory effect of ectopic expression of fgf3 on otx1. These data together suggest that Fgf17b plays a role in early embryonic patterning. We also demonstrate that fgf17b and fgf8 have stronger mesoderm inducting activity than fgf3, whereas fgf17b and fgf3 have stronger activity in posteriorizing the neuroectoderm than fgf8. Like fgf8, activation of fgf17b expression depends on Nodal signaling.

The development of semicircular canals in the inner ear: role of FGFs in sensory cristae

In the vertebrate inner ear, the ability to detect angular head movements lies in the three semicircular canals and their sensory tissues, the cristae. The molecular mechanisms underlying the formation of the three canals are largely unknown. Malformations of this vestibular apparatus found in zebrafish and mice usually involve both canals and cristae. Although there are examples of mutants with only defective canals, few mutants have normal canals without some prior sensory tissue specification, suggesting that the sensory tissues,cristae, might induce the formation of their non-sensory components, the semicircular canals. We fate-mapped the vertical canal pouch in chicken that gives rise to the anterior and posterior canals, using a fluorescent,lipophilic dye (DiI), and identified a canal genesis zone adjacent to each prospective crista that corresponds to the Bone morphogenetic protein 2 (Bmp2)-positive domain in the canal pouch. Using retroviruses or beads to increase Fibroblast Growth Factors (FGFs) for gain-of-function and beads soaked with the FGF inhibitor SU5402 for loss-of-function experiments,we show that FGFs in the crista promote canal development by upregulating Bmp2. We postulate that FGFs in the cristae induce a canal genesis zone by inducing/upregulating Bmp2 expression. Ectopic FGF treatments convert some of the cells in the canal pouch from the prospective common crus to a canal-like fate. Thus, we provide the first molecular evidence whereby sensory organs direct the development of the associated non-sensory components, the semicircular canals, in vertebrate inner ears.

Fgf3 signaling from the ventral diencephalon is required for early specification and subsequent survival of the zebrafish adenohypophysis

The pituitary gland consists of two major parts: the neurohypophysis, which is of neural origin; and the adenohypophysis, which is of non-neural ectodermal origin. Development of the adenohypophysis is governed by signaling proteins from the infundibulum, a ventral structure of the diencephalon that gives rise to the neurohypophysis. In mouse, the fibroblast growth factors Fgf8, Fgf10 and Fgf18 are thought to affect multiple processes of pituitary development: morphogenesis and patterning of the adenohypophyseal anlage; and survival, proliferation and differential specification of adenohypophyseal progenitor cells. Here, we investigate the role of Fgf3 during pituitary development in the zebrafish, analyzing lia/fgf3 null mutants. We show that Fgf3 signaling from the ventral diencephalon is required in a non-cell autonomous fashion to induce the expression of lim3,pit1 and other pituitary-specific genes in the underlying adenohypophyseal progenitor cells. Despite the absence of such early specification steps, fgf3 mutants continue to form a distinct pituitary anlage of normal size and shape, until adenohypophyseal cells die by apoptosis. We further show that Sonic Hedgehog (Shh) cannot rescue pituitary development, although it is able to induce adenohypophyseal cells in ectopic placodal regions of fgf3 mutants, indicating that Fgf3 does not act via Shh, and that Shh can act independently of Fgf3. In sum, our data suggest that Fgf3 signaling primarily promotes the transcriptional activation of genes regulating early specification steps of adenohypophyseal progenitor cells. This early specification seems to be essential for the subsequent survival of pituitary cells, but not for pituitary morphogenesis or pituitary cell proliferation.

Runx2 mediates FGF signaling from epithelium to mesenchyme during tooth morphogenesis

Runx2 (Cbfa1) is a runt domain transcription factor that is essential for bone development and tooth morphogenesis. Teeth form as ectodermal appendages and their development is regulated by interactions between the epithelium and mesenchyme. We have shown previously that Runx2 is expressed in the dental mesenchyme and regulated by FGF signals from the epithelium, and that tooth development arrests at late bud stage in Runx2 knockout mice [Development 126 (1999) 2911]. In the present study, we have continued to clarify the role of Runx2 in tooth development and searched for downstream targets of Runx2 by extensive in situ hybridization analysis. The expression of Fgf3 was downregulated in the mesenchyme of Runx2 mutant teeth. FGF-soaked beads failed to induce Fgf3 expression in Runx2 mutant dental mesenchyme whereas in wild-type mesenchyme they induced Fgf3 in all explants indicating a requirement of Runx2 for transduction of FGF signals. Fgf3 was absent also in cultured Runx2-/- calvarial cells and it was induced by overexpression of Runx2. Furthermore, Runx2 was downregulated in Msx1 mutant tooth germs, indicating that it functions in the dental mesenchyme between Msx1 and Fgf3. Shh expression was absent from the epithelial enamel knot in lower molars of Runx2 mutant and reduced in upper molars. However, other enamel knot marker genes were expressed normally in mutant upper molars, while reduced or missing in lower molars. These differences between mutant upper and lower molars may be explained by the substitution of Runx2 function by Runx3, another member of the runt gene family that was upregulated in upper but not lower molars of Runx2 mutants. Shh expression in mutant enamel knots was not rescued by FGFs in vitro, indicating that in addition to Fgf3, Runx2 regulates other mesenchymal genes required for early tooth morphogenesis. Also, exogenous FGF and SHH did not rescue the morphogenesis of Runx2 mutant molars. We conclude that Runx2 mediates the functions of epithelial FGF signals regulating Fgf3 expression in the dental mesenchyme and that Fgf3 may be a direct target gene of Runx2.

SOX7 and GATA-4 are competitive activators of Fgf-3 transcription

Fgf-3 is expressed in a dynamic and complex spatiotemporal pattern during mouse development. Previous studies identified GATA-4 as a transcription factor that binds the key regulatory element PS4A of the Fgf-3 promoter and stimulates transcription. Here we show that members of the SOX family of transcription factors also bind PS4A and differentially modulate transcription. At least five SOX genes, Sox2, Sox6, Sox7, Sox13, and Sox17, were expressed in F9 cells, and of these, Sox7 and Sox17 were dramatically induced in parallel with Fgf-3 following differentiation into parietal endoderm-like cells with retinoic acid and dibutyryl cAMP. Complexes could be detected on PS4A with SOX2, SOX7, and SOX17 by using nuclear extracts from differentiated F9 cells. However, only Sox7 expression markedly activated the Fgf-3 promoter in these cells. By contrast, SOX2 was a poor activator of Fgf-3 transcription, and when Sox2 was coexpressed with Gata4, it negatively modulated the strong activation mediated by GATA-4. More detailed analyses showed that SOX7 competes with GATA-4 for PS4A occupancy and to activate the Fgf-3 promoter. In situ hybridization analysis showed that Sox7 is co-expressed with Fgf-3 and Gata4 in the parietal endoderm of E7.5 mouse embryos. In culture, GATA-4-deficient embryonal stem cells were shown to express Fgf-3 upon differentiation into embryoid bodies, although at lower levels than were found in wild type embryonal stem cells. This Fgf-3 expression was virtually abolished when Sox7 expression was suppressed by RNA interference. These results show that SOX7 is a potent activator of Fgf-3 transcription.

High-throughput tissue microarray analysis of 11q13 gene amplification (CCND1, FGF3, FGF4, EMS1) in urinary bladder cancer

Gene amplification is a common mechanism for oncogene overexpression. High-level amplifications at 11q13 have been repeatedly found in bladder cancer by comparative genomic hybridization (CGH) and other techniques. Putative candidate oncogenes located in this region are CCND1 (PRAD1, bcl-1), EMS1, FGF3 (Int-2), and FGF4 (hst1, hstf1). To evaluate the involvement of these genes in bladder cancer, a tissue microarray (TMA) containing 2317 samples was screened by fluorescence in situ hybridization (FISH). The frequency of gains and amplifications of all genes increased significantly from stage pTa to pT1-4 and from low to high grade. In addition, amplification was associated with patient survival and progression of pT1 tumours. Among 123 tumours with amplifications, 68.3% showed amplification of all four genes; 19.5% amplification of CCND1, FGF4, and FGF3; and 0.8% co-amplification of FGF4, FGF3, and EMS1. Amplification of CCND1 alone was found in 9% of the tumours, while EMS1 alone was amplified in 1.6% and FGF4 in 0.8%. Overall, the amplification frequency decreased with increasing genomic distance from CCND1, suggesting that, among the genes examined, CCND1 is the major target gene in the 11q13 amplicon in bladder cancer.

Giant cell tumor of bone. The role of fibroblast growth factor 3 positive mesenchymal stem cells in its pathogenesis

OBJECTIVES

Giant cell tumor of bone is typified by massive infiltration of a bland neoplastic stroma by osteoclasts and monocyte progenitors. The current study aimed at evaluating the nature of the neoplastic cells and the mechanisms underlying the massive giant cell recruitment.

METHODS

Five different giant cell tumors were evaluated by immunohistochemistry, and explant cell cultures were established from the same tumors. Antigen expression profiles of both the tumors and the derived cultures were assessed. In order to determine if the mesenchymal cells are capable of differentiating into mature osteoblasts, retinoic acid was added to cell cultures and osteocalcin and alkaline phosphatase levels were measured. The proliferative effects of the mesenchymal cells on histiocyte-like cells were evaluated using the U-937 cell line.

RESULTS

A large stromal subpopulation expresses fibroblast growth factor receptor 3 (FGF-R3), indicating a mesenchymal origin of these cells. Few cells express bone- or cartilage-specific markers. Cell cultures are predominated by mesenchymal cells, as indicated by a strong staining by FGF R3. Retinoic acid induces osteoblastic differentiation, i.e. osteocalcin expression and alkaline phosphatase production. Conditioned medium of giant-cell-tumor-derived stromal cell cultures induces proliferation of U-937 cells, derived from histiocytic lymphoma. Papain digestion and dialysis of the conditioned media indicates the effector molecule to be a protein over 40 kD in size. The giant cell tumors as well as stromal cell cultures derived from giant cell tumors express osteoprotegerin ligand, the osteoclast activator.

CONCLUSIONS

The neoplastic stromal spindle-shaped subpopulation of cells in giant cell tumors are mesenchymal stem cells capable of inducing histiocyte proliferation. Retinoid acid is capable of inducing differentiation of the cells into mature osteoblasts. This should be further investigated in an in vivo model to ascertain whether induction of differentiation will prevent bone loss and retard tumor progression.

Nodal and Fgf pathways interact through a positive regulatory loop and synergize to maintain mesodermal cell populations

Interactions between Nodal/Activin and Fibroblast growth factor (Fgf)signalling pathways have long been thought to play an important role in mesoderm formation. However, the molecular and cellular processes underlying these interactions have remained elusive. Here, we address the epistatic relationships between Nodal and Fgf pathways during early embryogenesis in zebrafish. First, we find that Fgf signalling is required downstream of Nodal signals for inducing the Nodal co-factor One-eyed-pinhead (Oep). Thus, Fgf is likely to be involved in the amplification and propagation of Nodal signalling during early embryonic stages. This could account for the previously described ability of Fgf to render cells competent to respond to Nodal/Activin signals. In addition, overexpression data shows that Fgf8 and Fgf3 can take part in this process. Second, combining zygotic mutations in ace/fgf8 and oep disrupts mesoderm formation, a phenotype that is not produced by either mutation alone and is consistent with our model of an interdependence of Fgf8 and Nodal pathways through the genetic regulation of the Nodal co-factor Oep and the cell propagation of Nodal signalling. Moreover,mesodermal cell populations are affected differentially by double loss-of-function of Zoep;ace. Most of the dorsal mesoderm undergoes massive cell death by the end of gastrulation, in contrast to either single-mutant phenotype. However, some mesoderm cells are still able to undergo myogenic differentiation in the anterior trunk of Zoep;aceembryos, revealing a morphological transition at the level of somites 6-8. Further decreasing Oep levels by removing maternal oep products aggravates the mesodermal defects in double mutants by disrupting the fate of the entire mesoderm. Together, these results demonstrate synergy between oep and fgf8 that operates with regional differences and is involved in the induction, maintenance, movement and survival of mesodermal cell populations.

Requirements for FGF3 and FGF10 during inner ear formation

Members of the fibroblast growth factor (FGF) gene family control formation of the body plan and organogenesis in vertebrates. FGF3 is expressed in the developing hindbrain and has been shown to be involved in inner ear development of different vertebrate species, including zebrafish, Xenopus, chick and mouse. In the mouse, insertion of a neomycin resistance gene into the Fgf3 gene via homologous recombination results in severe developmental defects during differentiation of the otic vesicle. We have addressed the precise roles of FGF3 and other FGF family members during formation of the murine inner ear using both loss- and gain-of-function experiments. We generated a new mutant allele lacking the entire FGF3-coding region but surprisingly found no evidence for severe defects either during inner ear development or in the mature sensory organ, suggesting the functional involvement of other FGF family members during its formation. Ectopic expression of FGF10 in the developing hindbrain of transgenic mice leads to the formation of ectopic vesicles, expressing some otic marker genes and thus indicating a role for FGF10 during otic vesicle formation. Expression analysis of FGF10 during mouse embryogenesis reveals a highly dynamic pattern of expression in the developing hindbrain, partially overlapping with FGF3 expression and coinciding with formation of the inner ear. However, FGF10 mutant mice have been reported to display only mild defects during inner ear differentiation. We thus created double mutant mice for FGF3 and FGF10, which form severely reduced otic vesicles, suggesting redundant roles of these FGFs, acting in combination as neural signals for otic vesicle formation.