- Email: [email protected]
Murine FGF-12 and FGF-13: expression in embryonic nervous system, connective tissue and heart
ELSEVIER
Mechanisms of Development64 (1997) 31-39
Murine FGF-12 and FGF-13: expression in embryonic nervous system, connective tissue and heart H e l g e H a r t u n g a, B e n j a m i n F e l d m a n a, H e i k e L o v e c b, F r a n c o i s C o u l i e r b, D a n i e l B i r n b a u m b, M i t c h e l l G o l d f a r b a'* =Brookdale Centerfor Molecular Biology, Mount Sinai School of Medicine, 1 Gustave Levy Place, New York, NY 10029, USA bull9 1NSERM, 27 Boulevard Lei Roure, 13009 Marseille, France
Received 15 January 1997; revised version received 18 February 1997; accepted26 February 1997
Abstract
The molecular cloning of cDNAs encoding murine fibroblast growth factor-13 (FGF-13/FHF-2) and three isoforms of murine FGF-12 (FHF-1) is described. Like their highly conserved human counterparts, murine FGF-12 and FGF-13 are part of a distinct subfamily of FGFlike proteins characterized by a greater degree of amino acid sequence cross-homology and by conserved N-terminal domains which do not include secretion signal sequences. In addition to their expression in several adult tissues, both of these FGF genes are prominently and regionally expressed in midgestation mouse embryos, as revealed by in situ hybridization. Fgfl2 and fgfl3 RNAs were detected in developing central nervous system in cells outside the proliferating ependymal layer, and fgfl3 RNA was also found throughout the peripheral nervous system. Fgfl2 is expressed in developing soft connective tissue of the limb skeleton and in presumptive connective tissue linking vertebrae and fibs. Both FGF genes are also expressed in the myocardium of the heart, withfgfl2 RNA found only in the atrial chamber andfgfl3 RNA detected in both atrium and ventricle. On the basis of their novel structure and patterns of expression, FGF12 and FGF-13 are anticipated to perform embryonic functions distinct from other known FGF molecules. © 1997 Elsevier Science Ireland Ltd. Keywords: Fibroblast growth factor; Nervous system; Connective tissue; Heart
1. Introduction
Fibroblast growth factors (FGFs) are structurally related ligands originally discovered as stimulators of fibroblast or epithelial cell growth, but which are now known to exert a broad range of biological activities during vertebrate embryogenesis. FGFs are required for embryonic growth prior to gastrulation (Feldman et al., 1995), for correct execution of gastrulation (Amaya et al., 1991; Deng et al., 1994; Yamaguchi et al., 1994), for patterning of the central nervous system (Cox and Hemmati-Brivanlou, 1995; Kengaku and Okamoto, 1995; Lamb and Harland, 1995), and for the formation, growth, and shaping of many tissues and organs (reviewed in Goldfarb, 1996). These functions are
* Corresponding author.
achieved, in part, by the highly regulated expression of specific FGF genes during development. Ten well-characterized FGFs constitute a family on the basis of amino acid sequence homology within a core region of approximately 120 amino acid residues (Coulier et al., 1997). Eight of these FGFs (FGF-3-10) are constitutively secreted by cotranslational transport into the lumen of the endoplasmic reticulum and undergo glycosylation. Core homology imparts upon FGFs a common tertiary structure (Zhu et al., 1991) and the ability of FGFs to interact avidly with heparan glycosaminoglycans (Shing et al., 1984; Faham et al., 1996). FGFs complexed with heparin or heparan sulfate proteoglycan bind and activate cell surface FGF receptor tyrosine kinases (Shing et al., 1984; Rapraeger et al., 1991; Yayon et al., 1991; Faham et al., 1996). Four additional genes encoding proteins homologous to FGFs have recently been identified through searches of expressed sequence tag databases (Smallwood et al.,
0925-4773/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0925-4773(97)00042-7
32
H. Hartunget al. / Mechanismsof Development64 (1997)31-39
1996; Coulier et al., 1997; Verdier et al., 1997). On the basis of several criteria, these proteins, referred to as FGF-11 (or FGF homologous factor-3; FHF-3), FGF- 12 (FHF- 1), FGF13 (FHF-2), and FHF-4, define a distinct structural and, perhaps, functional sublineage within the FGF family. First, the four new FGFs bear greater core homology to one another than they do to other FGFs. Second, the new FGFs bear substantial homology to one another outside the core domain, a property not shared among any of the ten previously described FGFs. Third, none of these new FGFs bear N-terminal signal sequences for secretion through the vesicular pathway. Finally, while these new FGFs bind avidly to heparin, their ability to activate any of the known FGF receptors has not yet been demonstrated. These observations raise the possibility that, despite their homology to other FGFs, proteins in the new FGF sublineage function through distinct and as yet uncharacterized biochemical pathways. We report here on the characterization of murine fgfl2 andfgfl3 cDNA clones and on the expression of these two FGF genes in embryonic and adult tissues. The expression profiles for these genes are very different from those of other FGF genes, and suggest that FGF-12 and FGF-13 function in the development of the central and peripheral nervous systems, connective tissue of the skeleton, and the myocardia of the heart.
2. Results
2.1. Characterization of murine fgfl2 and fgfl3 cDNA clones The entire coding sequence of a humanfg,fl2 cDNA clone was used to make 32p-labelled hybridization probe for screening a neonatal mouse brain cDNA library at moderate stringency. Several hybridizing clones were purified and analyzed by PCR with FGF-specific oligonucleotide primers and by DNA sequencing. Two of these cDNA clones, p2-4A and p2-2B, derive from alternatively spliced forms of fgfl2 RNA differing in 5' sequence and predicted to encode FGF-12 isoforms with different N-termini. A third isoform of murine fgfl2 RNA was identified in survey of the NCBI dbEST sequence database. The hybridization screen also isolated cDNA clone p5-4 predicted to encode full-length murine FGF-13 protein. The predicted amino acid sequences for murine FGF-13 and different isoforms of murine FGF-12 are shown in Fig. 1A, along with those for human FGF-1, -5, -9, -12, and -13. One isoform of murine FGF-12, here termed mFGF-12A, is 243 amino acids in length and is 99% identical to the previously described human FHF-1 (hFGF-12A). FGF12A contains an FGF core homology domain along with 76 N-terminal and 42 C-terminal residues. A second isoform of murine FGF-12, termed mFGF-12B, is only 181 amino acids in length, owing to a different and much shorter
N-terminal extension. The 12B isoform is also conserved in humans, and is represented by several dbEST sequence IDs. Only a partial sequence of an mFGF-12C isoform is available. Since this isoform does not yet have a characterized human counterpart, the likelihood of the transcript producing functional FGF-12C protein cannot be assessed. That these 5' sequence differences among fgfl2 cDNA clones result from alternative splicing was confirmed by the sequencing of a murine genomic fgfl2 clone within the region spanning coding exon 2 (data not shown). Murine FGF- 13 has a predicted length of 244 amino acid residues and is 99% identical to human FGF-13 (FHF-2). mFGF-13 and mFGF- 12A are 78% identical within the FGF core homology region (Fig. 1B), 59% identical within their N-terminal extensions, and 45% identical within their Cterminal extensions. Like their human counterparts, murine FGF- 12A, FGF- 12B, and FGF- 13 lack N-terminal secretion signal sequences. By contrast, all of the FGFs are linked by weaker core sequence homology (Fig. 1B).
2.2. Expression of FGF-12 and FGF-13 RNA within the embryonic nervous system We have analyzed the major sites of fgfl2 and fgfl3 expression during mouse embryogenesis by in situ hybridization using 35S-labelled antisense riboprobes. PCR-derived fragments of murine fgfl2 and fgfl3 cDNAs were used as templates for synthesis of riboprobes which were hybridized to deparaffinized sections of mouse embryos ranging in development from 9.5 to 16.5 days post coitum (E9.5E16.5). Overall expression of both genes became evident at El0.5, was much stronger at E12.5 and E14.5, and declined at most sites by E16.5. Within the central nervous system, fgfl 2 and fgH3 expression appears in differentiating cells which have migrated out from the proliferating ependymal layer of neuroepithelium. In the E10.5 spinal cord, predominant expression is restricted ventrally where differentiating cells first become evident (Fig. 2A-D). At this stage, the domains of fgfl2 and fgfl3 expression are overlapping, but not identical. Whereas fgfl2 is expressed bilaterally in the ventromedial region of spinal cord including the floor plate (Fig. 2A,B), fgfl3 is expressed in the ventrolateral horns which contain motoneurons (Fig. 2C,D). At this developmental stage, FGF expression also appears in brain outside the ependymal zones. Within the El0.5 mesencephalic region, fgfl2 and fgfl3 expression is prominent in thin bands of cells just outside the ependymal layer (Fig. 2E-G). By E12.5, fgfl2 and fgfl3 are both strongly expressed outside the ependymal zone in the spinal cord (Fig. 2H,I) and brain (data not shown) along their entire rostrocaudal and dorsoventral axes. Widespread CNS expression of both FGFs persists at E14.5 (Fig. 2L,M), with restricted regions of intense expression evident (e.g. Fig. 2J,K). fgfl3 expression declines, but is still detectable, in E16.5 CNS, while fgfl2 CNS expression is generally below detection (data not
33
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
A) mFGFI2A hFGFI2A mFGFI2B hFGFI2B mFGFI2C mFGFI3 hFGFI3 hFGFI hFGF5 hFGF9
MAAAIASSLIRQKRQARESNSDRVSASKRRSSPSKDGRSLCDRHVLGVFSKVRFCSGRKRPVRRRPEPQLKGIVTR **************************************************************************** MESK********** MESK********** .... * G R S L * S S D S P S * * * * * * * ~ * * *-*****************--***************-**********************-************** *-*****************--***************-*****--***************-************** MAEGEITTFTALTEKFNLPPGNYKKPKL MSLSFLLLLFFSHLILSAWAHG•KRLAPKGQPGPAATDRNPIGSSSRQSSSSAMSSSSASSSPAASLGSQGSGLEQSSFQWS•SGRRTGS MAPLGEVGNYFGVQDAVPFGNVPVLPVDSPVLLSDHLGQSEAGGLPRGPAVTDLDHLKGILRRRQ
mFGFI2A hFGFI2A mFGFI3 hFGFI3
LFSQQG••YFLQMH•DGTIDGTKDENSD•YTL•FNL•pV-GLRwAIQGVKASLYVA••NGEGYLYSSDVFT••CKFKESVFENYYVIYSSTLYRQQ•SGRAW ******--********************_***_******_************************************************************** ******••*•**•*•**********•*••*********************••***•***•*****•*••*****************•***•*********** ******--********************-***-******-**************************************************************
hFGFI hFGF5 hFGF9
*YCS•*-GH**RIL****•***R*R-**QHI•LQLSAESA*E--*Y*•*TETGQ*L**DT**L**G*QTPNE**L*L*•LE**H*NT*T*KKHAEKN *•CRV*IGFH**IY***K"v-N*SHEA*•L--SVLEIFAVSQ*I--*G*•**FS•KFL**SKK*K*HA*AK**DD***R*RFQ**S*NT*A*AIH*TEKT**•* *YCRT*--FH*EIF*N***•**RKDH*•FGI*•*IS*A*-**--*S*R**DSG**LG**EK*E**G*EKL*•**V*•*•FE**W*NT***•**KHVDT**RY
m F G F 12 A hFGFI2A mFGFI3 hFGFI3 hF G F 1 hFGF5 hFG F 9
FLGLNKEGPIMKG- -NRVKKTKPSSHFVPKPIEVCMYREPSLHEIGEKQGRSRKSSGTPT *************--********************************************* **
* * * * * * E* * **--***
*****
**
* * * * * * E* * **--***
*****A~.*
~.
* ***
* *********
* **
******___******
***
* *********
* **
******---*******
*V* * K * N * S C K R * - - P * T H Y G Q K A I L * L *
........ MNGGK-VVNQDST ........ *****-****** * * * ****,****** * ************
* * ********
*
* ********
*
76 76 14 14 71 71 28 90 65
173 173 168 168 .... *
122 188 162
243 243 244 244 153
L*VS
YVA* **R*KAKR*CSP***PQHI*T**L*RFKQSEQPELSFTVTVP**KNPPSPIKSKIPLSAPRKNTNSVKYRLKFRFG Y V A * * * D * T P R E * - - T * T * R H O K F T * *~ * R P V D P D K V P E L Y K D I L S Q S
268 208
B) FGF-1 100
FGF-5 40 100
FGF-9 41 45 100
FGF-12 35 35 .42 100
FGF-13 36 38 45 78 100
FGF-1 FGF-5 FGF-9 FGF-12 FGF-13
Fig. 1. FGF protein sequence alignment. (A) The predicted amino acid sequences for human and mouse FGFs are aligned, with the FGF core homology domain underlined, and with residues identical to those in murine FGF-12A indicated as '*', gaps indicated as '-', and the incomplete N-terminus of mFGF12C indicated as '...'. The sequence of murine FGF-12A (mFGF-12A) is derived from cDNA clone p2-4A (this study); human FGF-12A (hFGF-12A) (Smallwood et al., 1996); mFGF- 12B is from NCBI dbEST ID 299539 (Genbank W05845); hFGF-12B is from NCBI dbEST IDs 279066 (Genbank H19128) and 289370 (Genbank H28811) and our corrections by direct sequencing (Coulier et al., 1997); mFGF-12C from cDNA clone p2-2B (this study); mFGF-13 from cDNA clone p5-4 (this study); hFGF-13 (Smallwood et al., 1996) and correction to the N-terminal sequence based upon EMBL Z65100; hFGF-1 (Gimenez-Gallego et al., 1986), hFGF-5 (Haub and Goldfarb, 1991), and hFGF-9 (Miyamoto et al., 1993). (B) Pairwise percent amino acid sequence homology within the FGF core domain. FGF-12 and FGF-13 core domains are strongly homologous to each other and bear homology to other FGFs within the range typical for the FGF family.
shown), except in the retina, where strongfgfl2 expression persists in the retinal ganglion layer (Fig. 2N-P). fgf13 is prominently expressed from El0.5 to E16.5 throughout the peripheral nervous system, including expression in dorsal root ganglia (Fig. 2C,D,J,K and 3C), all cranial ganglia (Fig. 2E,F and 3D,E), and ganglia of the enteric nervous system (Fig. 3A,B). Fgf12 shows little or no expression in the peripheral nervous system.
2.3. Expression of FGF-12 in developing connective tissue of the skeleton
Fgfl2 expression in the cartilagenous skeleton suggests a role for this factor in the development of connective tissue. The condensations within the handplate of the fore-limb express fgf12 as early as El0.5 (Fig. 4A,B). By E12.5, fgf12 expression becomes restricted to the external portions
of phalangeal condensations (Fig. 4C,D), while the interiors of the condensations commence chondrification, as determined morphologically and by the onset of ul(II) collagen (col2al) expression (Andrikopoulos et al., 1992) (Fig. 4E,F). Similarly, fgf12 is expressed towards the exterior of the tibial condensation of the hindlimb and within a zone spanning the prospective knee joint (Fig. 4G,H), while the centers of the tibial and femoral condensations, but not the joint, express col2al (Fig. 4H,I). By E14.5, strong fg,f12 expression in the hindlimb is clearly segregated to the dense non-chondrified mesenchyme which constitutes the skeletal connective tissue (Fig. 4J,K). These new FGFs are also expressed within the trunk at presumed sites of future skeletal connective tissue. Fgf12 expression is prominent in E12.5 mesenchyme adjacent to vertebral neural arch chondrifications (Fig. 4L,M) and in mesenchyme separating vertebrae from the base of rib shafts
34
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
(Fig. 4N-Q). Skeletal expression offgfl2 is not detectable by embryonic day 16, and fgf13 is not coexpressed at any sites of developing connective tissue.
2.4. FGF-12 and FGF13 expression in developing heart Fgf12 and fgf13 are expressed in the myocardium of the developing heart from El0.5 through E14.5. Sagittal sections through the heart of El0.5 and El2.5 embryos shows fgfl2 expression restricted to the myocardium of the atrial chamber (Fig. 5A,B,E). By contrast, fgf13 is expressed in both atrial and ventricular myocardia, but is not expressed in endocardial cushions (Fig. 5C,D). 2.5. Other sites of FGF-12 and FGF-13 embryonic expression 2.5.1. Axial skeleton When primitive sclerotomal tissue becomes rostrocaudally organized into alternating bands of looser and denser mesenchyme, fgf12 is expressed in the bands of loose mesenchyme. Upon further reorganization of the sclerotomal tissue into chondrogenic vertebral body masses and intervertebral tissue, fgf12 expression is downregulated. This is illustrated in the caudal region at embryonic day 14.5 (Fig. 6A,B), where fgf12 is strongly expressed in the bands of looser sclerotomal mesenchyme in the tail, but expression has diminished in more rostral (sacral) segments which have undergone reorganization. 2.5.2. Craniofacial mesenchyme Fgf12 is expressed in regions of craniofacial mesenchyme between El0.5 and E14.5. As one example, basiooccipital mesenchyme just beneath the brain bears segments of fgf12 expression at El0.5 (Fig. 2G) and E12.5 (Fig. 6E,F).
Fig. 2. Expression of murine fgfl2 and fgfl3 RNAs in embryonic central nervous system. Deparaffinized embryo sections were hybridized with 35Slabelled antisense riboprobes for murinefgcf12 orfgfl3, exposed to photographic emulsion, stained, and viewed under bright-field (A,C,E,H,J,N) or dark-field (B.D,F,G+I,K-M,O,P) illumination. (A,B)fgff12, El0.5, spinal cord, transverse section, expression in floor plate and ventromedial spinal cord; (C,D)fgf13, El0.5, spinal cord, transverse, expression in ventral horns and dorsal root ganglia; (E,F)fgf13, El0.5, head, longitudinal. expression in trigeminal ganglia and midbrain outside ependymal layer (arrowheads); (G)fgoC12. El0.5, head, longitudinal, expression outside ependymal layer (arrowheads) and in basiooccipital mesenchyme (arrows); (H,I) fgfl2, E12.5, spinal cord, transverse, expression in floor plate and outside ependymal layer; (J,K)fgcf13, E 14.5, spinal cord, transverse, broad expression with regionally strong expression; (L)fgf12, E14.5, whole embryo, sagittal, expression throughout brain and spinal cord; (M) FGF13, E14.5, whole embryo, sagittal, expression throughout brain and spinal cord; (N,O)fgf12, E16.5, eye, sagittal, strong expression in retinal ganglion layer and weak expression in lens; (P)fgf13, E16.5, eye, sagittal, no detectable expression. Spinal cord (sc), ependymal layer (el, floor plate (f), ventral horn (vh/, dorsal root ganglion (d), midbrain (rob), trigeminal ganglion (t), oropharynx (p). vertebral body (v), neuroepithelium (n), retinal ganglion layer (g), lens (le).
......
:
+
.
:~I+
....
•
+
.
H. Hartunget al. /Mechanismsof Development64 (1997)31-39
35
Fig. 3. Fgf13RNA in E14.5 peripheral nervous system. Bright field (A,C,D) and dark field (B,E) offgfl3 hybridization to sections of E14.5 embryo. (A,B) Intestine, cross-section,expressionin enteric ganglia situated between inner circular and outer longitudinal muscle layers; (C) strong expression in dorsal root ganglia; (D,E) head, parasagittal section, strong expression in fifth, seventh, eighth, and ninth cranial ganglia. Luminal epithelium (e), mesenchymallayer (m), inner muscular layer (i), outer muscular layer (o), dorsal root ganglia (d), vertebral body (v), canals of inner ear (c), cranial ganglia (5,7,8,9), brain (b).
2.5.3. Nasal epithelium Fgf12 is expressed in nasal olfactory epithelium at E12.5 (Fig. 6C,D) and E14.5 (data not shown).
2.5.4. Kidney Fgf12 andfgf13 are expressed in non-overlapping regions of the developing kidney. Whereas fgf12 expression is restricted to glomeruli (Fig. 6E,F),fgfl3 is expressed in patches of loose mesenchyme surrounding the collecting duct and near glomeruli (Fig. 6G,H).
16 tissues analyzed. By contrast, fgf12 R N A species were seen in heart, brain, prostate, ovaries, and intestine, but not in other tissues. The 6.0, 2.6, and 1.2 kb species offgfl2 RNA seen in varying relative abundance may result from alternative pre-mRNA splicing, as described earlier in the analysis of fgf12 c D N A clones. These findings stand in sharp contrast to other FGF genes, whose expression in adult tissues is highly restricted or completely absent (Dickson et al., 1989; Curatola and Basilico, 1990; de Lapeyriere et al., 1990).
2.6. Expression in adult tissues 3. Discussion The expression offgfl2 andfgfl3 genes in adult human tissues was analyzed by Northern blot hybridization with 32p-labelled human FGF cDNA fragments. As shown in Fig. 7, a 2.1 kb fgf13 RNA transcript was detectable in all
Fgfl2 andfgfl3 genes are dynamically expressed in midgestation mouse embryos at sites different from those described for other previously characterized FGFs. While
36
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
Fig. 4. Fgfl2 RNA in developing skeletal connective tissue. Hybridizations with fgfl2 or col2al viewed in bright (A,C,E,H,J,L,N,P) or dark field (B,D,F,G,I,K,M,O,Q). (A,B)fgfl2, El0.5, handplate, cross-section, expression condensations; (C,D)fgf12, E12.5, handplate, cross-section, expression in exterior of phalangeal condensations; (E,F) col2al, section adjacent to (C,D), expression in interior of condensations: (G)fgf12, E12.5, hindlimb, oblique, expression in exterior of tibial condensation and in knee joint: (H,I) col2al, section adjacent to (G), expression within tibial and femural condensations; (J,K) fgf12, E14.5, hindlimb, sagittal, expression in dense connective tissue mesenchyme and virtual exclusion from cartilagenous condensations (*); (L,M)fgf12, E12.5, cervical region, expression in mesenchyme near vertebral neural arches; (N,O) parasagittal and (P,Q) longitudinal sections, E12.5,fgf12, expression in mesenchyme between chondrifications of vertebral bodies and bases of rib shafts. Forelimb (ft), hindlimb (hi), prospective tibia (ti), femur (fe), knee joint (j), vertebral neural arch (n), vertebral body (v), rib shaft (r), spinal cord (sc).
H. Hartung et aL /Mechanisms of Development 64 (1997) 31-39
37
fgf3 and fgf8 are transiently expressed within restricted domains of brain neuroepithelium (Wilkinson et al., 1989; Crossley and Martin, 1995), fgfl2 and fgfl3 are, so far, unique among FGFs in their prominent and widespread expression throughout the embryonic CNS and PNS. Similarly, expression offgfl2 andfgfl3 in developing heart and connective tissue is, so far, unique among FGFs. These observations suggest several novel embryonic functions for these newly identified FGFs. The use of gene targetting in mice to disruptfgfl2 orfgfl3 expression and of transgenic methods to promote regional ectopicfg(12 orfgfl3 expression should reveal certain embryonic requirements for these factors. Several enigmatic properties of FGF-12 and FGF-13, along with closely related FGF-11 (FHF-3) and FHF-4, complicate speculation regarding their biological and biochemical functions. These FGFs form a distinct sublineage of FGFs characterized by strong homology within the FGF core domain along with substantial homology outside of the core. While recombinant FGF-12 and FGF-13 proteins share with other FGFs the ability to bind avidly to heparin-agarose, FGF-12 and FGF-13 cannot activate many of the known FGF receptor tyrosine kinases (H. Hartung and M. Goldfarb, unpublished data). Additionally, FGF-12 and FGF-13 fail to mimic the ability of other FGFs to induce mesoderm in animal caps excised from blastula-stage frog embryos (A. Suzuki, H. Hartung, M. Goldfarb and A. Hemmati-Brivanlou, unpublished data). Furthermore, no members of the new FGF subfamily bear N-terminal signal sequences for secretion, and FGF-12/ FHF-1 protein has been shown to accumulate intracellularly in vivo (Smallwood et al., 1996). It is possible that these new FGFs activate novel cell surface receptors or interact non-productively (antagonistically) with known FGF receptors. Alternatively, the FGF core homology domain may represent a functionally versatile structural module utilized by FGF-12 and FGF-13 for novel molecular interactions within the cell.
4. Materials and methods
4.1. Cloning and sequencing murine FGF-12 and FGF-13 cDNAs
Fig. 5. Fg~12 and fgfl3 RNAs in embryonic heart. Detection of fgfl2 (A,B,E) and fgfl3 (C,D) RNAs in parasagittal sections of El0.5 (A,B) or E12.5 (C-E) heart viewed under bright field (A,C) or dark field (B,D,E). Atrial myocardiurn (a), ventricular myocardium (v), endocardial cushions (e), and liver (li).
The complete coding sequence of a human fgfl2 cDNA clone (Coulier et al., 1997) (American Type Culture Collection, IMAGE Consortium) was amplified by polymerase chain reaction (PCR), 32p-labelled with random hexamer primers and DNA polymerase I Klenow fragment, and used to screen a whole mouse E17.5 cDNA lamda ZAP library (Stratagene) by filter transfer plaque hybridization. Purified positive clones were converted to plasmids by in vivo excision and their identities determined by PCR with FGF-specific primers followed by dideoxynucleotide sequencing. The coding sequences of murine fgfl3 cDNA
H. Hartung et al./ Mechanisms of Development 64 (1997) 31-39
38
i
C(;:;
• ::•2
Fig. 6. Other sites offgfl2 andfgf13 embryonic expression. Bright field (A,C,E,G,I) and dark field (B,D,F,H,J) illumination. (A,B)fgfl2, E14.5, caudal region, mid-sagittal, segmental expression in the loose mesenchyme bands (brackets) of tail sclerotomes but diminishing in segments undergoing vertebral differentiation; (C,D)fgf12, E12.5, snout, parasagittal, expression in nasal olfactory epithelium; (E,F)fgfl2, E12.5, head, longitudinal, expression in basiooccipital region above tongue; (G,H)fgfl2 and (I,J)fgfl3, E14.5, kidney, parasagittai, withfgf12 RNA in glomeruli andfgfl3 RNA in mesenchyme. Sacral vertebral body (v), brain vesicle (bv), nasal epithelium (n), snout (sn), tongue (t), basiooccipital region (bo), kidney collecting duct (d), glomeruli (g).
h
b
pl lu Ii sm k pa
~
$p th IX te o si c phi
kb
$p th pr te o si ¢ phi
m mmm
FGF12
e h
1
b
131 lu Ii sm k pa . . . . . .
~
O.s..-~:~:~
: i ii~ ~
~
FGF13
Fig. 7. Northern blot analysis offgfl2 andfgfl3 expression in adult human tissues. Human polyA + RNA from different tissues were electrophoresed through parafbrmaldehyde agarose gels, transferred to nylon membranes, and hybridized with 32p-labelled fragments of humanfgfl2 orfg~13 cDNAs. Washed filters were autoradiographed for 3-7 days. Heart (h), brain (b), placenta (pl), lung (lu), liver (li) skeletal muscle (sm), kidney (k), pancreas (p), spleen (sp), thymus (th), prostate (pr), testis (te), ovary (o), small intestine (si), colon (c), and peripheral blood lymphocyte (PBL). The sizes of molecular weight markers are indicated.
1t. Hartung et al. / Mechanisms of Development 64 (1997) 31-39 clone p5-4, mfgfl2A c D N A clone 2-4A, and the alternative N-terminal coding sequence of mfgf12C c D N A clone 2-2B have been deposited in Genbank.
4.2. In situ hybridization Segments ofmurinefgH2,fgfl3, and col2al cDNAs were used as templates for antisense-strand 35S-labelled riboprobe synthesis with T7 R N A polymerase. Templates were as follows: mfgf12, 1.0 kb whole c D N A insert of plasmid p2-2B; mfgf13, 1.6 kb fragment of p5-4 c D N A insert, containing 200 bp C-terminal coding sequence and 1.4 kb 3' untranslated sequence; col2al, 450 bp fragment of 3' untranslated sequence (Andrikopoulos et al., 1992). Hybridization of probes to deparaffinized sections of mouse embryos was performed by the method of Wilkinson (Wilkinson and Nieto, 1993), with modification of the hybridization solution by inclusion of 20 m M sodium phosphate (pH 7.5). After hybridization, slides were dipped in undiluted Kodak NTB2 emulsion, exposed 10-21 days (except 5 days for col2al probe), developed and stained with hematoxylin/eosinY. Bright- and dark-field images viewed on Nikon Microphot and S M Z - U microscopes were captured with a Kodak DCS410 digital camera.
4.3. Northern blot hybridization Human multiple tissue Northern blots (MTN and MTNII, Clontech Laboratories) were hybridized and washed under manufacturer's conditions using fragments o f F G F cDNAs 32p-labelled by random hexamer priming as probes. Probes were as follows: fgf12, 0.9 kb EcoRUNotI insert f r o r n plasmid yr45d03 (Genbank ID H62672); fgfl3, 0.75 kb PstI fragment from plasmid yg96e06 (Genbank ID R59031). Washed blots were exposed to Kodak X A R film for 3 - 7 days.
Acknowledgements W e thank Sam Bouyain for excellent assistance with D N A sequencing. M.G. is a research fellow of the Lucille P. Markey Charitable Trust.
39
References Amaya, E., Musci, T.J. and Kirschner, M.W. (1991) Cell 66, 257-270. Andrikopoulos,K., Suzuki, H.R., Solursh, M. and RamireL F. (1992) Dev. Dyn. 195, 113-120. Coulier, F., Pontarotti, P., Roubin, R., Hartung, H., Goldfarb M. and Birnbaum, D. (1997) J. Mol. Evol. 44, 43-56. Cox, W.G. and Hemmati-Brivanlou, A. (1995) Development 121, 43494358. Crossley, P.H. and Martin, G.R. (1995) Development 121, 439-451. Curatola, A.M. and Basilico, C. (1990) Mol. Cell. Biol. 10, 2475-2484. de Lapeyriere, O., Rosnet, O., Benharroch, D., Raybaud, F., Marchetto, S., Planche, J., Galland, F., Mattei, M.-G., Copeland, N.G., Jenkins, N.A., Coulier, F. and Birnbaum, D. (1990) Oncogene 5, 823-831. Deng, C.-X., Wynshaw-Boris, A., Shen, M.M., Daugherty, C., Omitz, D.M. and Leder, P. (1994) Genes Dev. 8, 3045-3057. Dickson, C., Deed, R., Dixon, M. and Peters, G. (1989) Prog. Growth Factor Res. 1. 123-132. Faham, S., Hileman, R.E., Fromm, J.R., Linhardt, R.J. and Rees, D.C. (1996) Science 271, 1116-1120. Feldman, B., Pouymeirou, W., Papaioannou, V.E., DeChiara, T. and Goldfarb, M. (1995) Science 267, 246-249. Gimenez-Gallego, G., Conn, G., Hatcher, V.B. and Thomas, K.A. (1986) Biochem. Biophys. Res. Commun. 138, 611-617. Goldfarb, M. (1996) Cytokine Growth Factor Rev. 7, 311-325. Haub, O. and Goldfarb, M. (1991) Development 112, 397-406. Kengaku, M. and Okamoto, H. (1995) Development 121, 3121-3130. Lamb, T.M. and Harland, R.M. (1995) Development 121. 3627-3636. Miyamoto, M., Naruo, K.-I., Seko, C., Matsumoto, S., Kondo, T. and Kurokawa, T. (1993) Mol. Cell. Biol. 13, 4251-4259. Rapraeger, A.C., Krufka, A. and Olwin, B.B. (1991) Science 252, 17051708. Shing, Y., Folkman, J., Sullivan, R., Butterfield, C., Murray, J. and Klagsbrun, M. (1984) Science 223, 1296-1299. Smallwood, P.M.. Munoz-Sanjuan, I., Tong, P., Macke, J.P., Hendry, S.H.C., Gilbert, D.J., Copeland, N.G., Jenkins, N.A. and Nathans, J. (1996) Proc. Natl. Acad. Sci. USA 93, 9850-9857. Verdier, A.-S., Mattei, M.-G., Lovec, H., Hartung, H., Goldfarb, M., Bimbaum, D. and Coulier, F. (1997) Genomics 40, 151-154. Wilkinson, D.G., Bhatt, S. and McMahon, A.P. (1989) Development 105, 131-136. Wilkinson. D.G. and Nieto, M.A. (1993) Methods Enzymol. 225, 361373. Yamaguchi. T.P., Harpal, K., Henkemeyer, M. and Rossant, J. (1994) Genes Dev. 8, 3032-3044. Yayon, A., Klagsbrun, M.. Esko, J.D., Leder, P. and Omitz, D.M. (1991) Cell 64, 841-848. Zhu, X., Komiya, H., Chirino, A., Faham, S., Fox, G.M., Arakawa, T., Hsu, B.T. and Rees, D.C. (1991) Science 251.90-93.
Mechanisms of Development64 (1997) 31-39
Murine FGF-12 and FGF-13: expression in embryonic nervous system, connective tissue and heart H e l g e H a r t u n g a, B e n j a m i n F e l d m a n a, H e i k e L o v e c b, F r a n c o i s C o u l i e r b, D a n i e l B i r n b a u m b, M i t c h e l l G o l d f a r b a'* =Brookdale Centerfor Molecular Biology, Mount Sinai School of Medicine, 1 Gustave Levy Place, New York, NY 10029, USA bull9 1NSERM, 27 Boulevard Lei Roure, 13009 Marseille, France
Received 15 January 1997; revised version received 18 February 1997; accepted26 February 1997
Abstract
The molecular cloning of cDNAs encoding murine fibroblast growth factor-13 (FGF-13/FHF-2) and three isoforms of murine FGF-12 (FHF-1) is described. Like their highly conserved human counterparts, murine FGF-12 and FGF-13 are part of a distinct subfamily of FGFlike proteins characterized by a greater degree of amino acid sequence cross-homology and by conserved N-terminal domains which do not include secretion signal sequences. In addition to their expression in several adult tissues, both of these FGF genes are prominently and regionally expressed in midgestation mouse embryos, as revealed by in situ hybridization. Fgfl2 and fgfl3 RNAs were detected in developing central nervous system in cells outside the proliferating ependymal layer, and fgfl3 RNA was also found throughout the peripheral nervous system. Fgfl2 is expressed in developing soft connective tissue of the limb skeleton and in presumptive connective tissue linking vertebrae and fibs. Both FGF genes are also expressed in the myocardium of the heart, withfgfl2 RNA found only in the atrial chamber andfgfl3 RNA detected in both atrium and ventricle. On the basis of their novel structure and patterns of expression, FGF12 and FGF-13 are anticipated to perform embryonic functions distinct from other known FGF molecules. © 1997 Elsevier Science Ireland Ltd. Keywords: Fibroblast growth factor; Nervous system; Connective tissue; Heart
1. Introduction
Fibroblast growth factors (FGFs) are structurally related ligands originally discovered as stimulators of fibroblast or epithelial cell growth, but which are now known to exert a broad range of biological activities during vertebrate embryogenesis. FGFs are required for embryonic growth prior to gastrulation (Feldman et al., 1995), for correct execution of gastrulation (Amaya et al., 1991; Deng et al., 1994; Yamaguchi et al., 1994), for patterning of the central nervous system (Cox and Hemmati-Brivanlou, 1995; Kengaku and Okamoto, 1995; Lamb and Harland, 1995), and for the formation, growth, and shaping of many tissues and organs (reviewed in Goldfarb, 1996). These functions are
* Corresponding author.
achieved, in part, by the highly regulated expression of specific FGF genes during development. Ten well-characterized FGFs constitute a family on the basis of amino acid sequence homology within a core region of approximately 120 amino acid residues (Coulier et al., 1997). Eight of these FGFs (FGF-3-10) are constitutively secreted by cotranslational transport into the lumen of the endoplasmic reticulum and undergo glycosylation. Core homology imparts upon FGFs a common tertiary structure (Zhu et al., 1991) and the ability of FGFs to interact avidly with heparan glycosaminoglycans (Shing et al., 1984; Faham et al., 1996). FGFs complexed with heparin or heparan sulfate proteoglycan bind and activate cell surface FGF receptor tyrosine kinases (Shing et al., 1984; Rapraeger et al., 1991; Yayon et al., 1991; Faham et al., 1996). Four additional genes encoding proteins homologous to FGFs have recently been identified through searches of expressed sequence tag databases (Smallwood et al.,
0925-4773/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0925-4773(97)00042-7
32
H. Hartunget al. / Mechanismsof Development64 (1997)31-39
1996; Coulier et al., 1997; Verdier et al., 1997). On the basis of several criteria, these proteins, referred to as FGF-11 (or FGF homologous factor-3; FHF-3), FGF- 12 (FHF- 1), FGF13 (FHF-2), and FHF-4, define a distinct structural and, perhaps, functional sublineage within the FGF family. First, the four new FGFs bear greater core homology to one another than they do to other FGFs. Second, the new FGFs bear substantial homology to one another outside the core domain, a property not shared among any of the ten previously described FGFs. Third, none of these new FGFs bear N-terminal signal sequences for secretion through the vesicular pathway. Finally, while these new FGFs bind avidly to heparin, their ability to activate any of the known FGF receptors has not yet been demonstrated. These observations raise the possibility that, despite their homology to other FGFs, proteins in the new FGF sublineage function through distinct and as yet uncharacterized biochemical pathways. We report here on the characterization of murine fgfl2 andfgfl3 cDNA clones and on the expression of these two FGF genes in embryonic and adult tissues. The expression profiles for these genes are very different from those of other FGF genes, and suggest that FGF-12 and FGF-13 function in the development of the central and peripheral nervous systems, connective tissue of the skeleton, and the myocardia of the heart.
2. Results
2.1. Characterization of murine fgfl2 and fgfl3 cDNA clones The entire coding sequence of a humanfg,fl2 cDNA clone was used to make 32p-labelled hybridization probe for screening a neonatal mouse brain cDNA library at moderate stringency. Several hybridizing clones were purified and analyzed by PCR with FGF-specific oligonucleotide primers and by DNA sequencing. Two of these cDNA clones, p2-4A and p2-2B, derive from alternatively spliced forms of fgfl2 RNA differing in 5' sequence and predicted to encode FGF-12 isoforms with different N-termini. A third isoform of murine fgfl2 RNA was identified in survey of the NCBI dbEST sequence database. The hybridization screen also isolated cDNA clone p5-4 predicted to encode full-length murine FGF-13 protein. The predicted amino acid sequences for murine FGF-13 and different isoforms of murine FGF-12 are shown in Fig. 1A, along with those for human FGF-1, -5, -9, -12, and -13. One isoform of murine FGF-12, here termed mFGF-12A, is 243 amino acids in length and is 99% identical to the previously described human FHF-1 (hFGF-12A). FGF12A contains an FGF core homology domain along with 76 N-terminal and 42 C-terminal residues. A second isoform of murine FGF-12, termed mFGF-12B, is only 181 amino acids in length, owing to a different and much shorter
N-terminal extension. The 12B isoform is also conserved in humans, and is represented by several dbEST sequence IDs. Only a partial sequence of an mFGF-12C isoform is available. Since this isoform does not yet have a characterized human counterpart, the likelihood of the transcript producing functional FGF-12C protein cannot be assessed. That these 5' sequence differences among fgfl2 cDNA clones result from alternative splicing was confirmed by the sequencing of a murine genomic fgfl2 clone within the region spanning coding exon 2 (data not shown). Murine FGF- 13 has a predicted length of 244 amino acid residues and is 99% identical to human FGF-13 (FHF-2). mFGF-13 and mFGF- 12A are 78% identical within the FGF core homology region (Fig. 1B), 59% identical within their N-terminal extensions, and 45% identical within their Cterminal extensions. Like their human counterparts, murine FGF- 12A, FGF- 12B, and FGF- 13 lack N-terminal secretion signal sequences. By contrast, all of the FGFs are linked by weaker core sequence homology (Fig. 1B).
2.2. Expression of FGF-12 and FGF-13 RNA within the embryonic nervous system We have analyzed the major sites of fgfl2 and fgfl3 expression during mouse embryogenesis by in situ hybridization using 35S-labelled antisense riboprobes. PCR-derived fragments of murine fgfl2 and fgfl3 cDNAs were used as templates for synthesis of riboprobes which were hybridized to deparaffinized sections of mouse embryos ranging in development from 9.5 to 16.5 days post coitum (E9.5E16.5). Overall expression of both genes became evident at El0.5, was much stronger at E12.5 and E14.5, and declined at most sites by E16.5. Within the central nervous system, fgfl 2 and fgH3 expression appears in differentiating cells which have migrated out from the proliferating ependymal layer of neuroepithelium. In the E10.5 spinal cord, predominant expression is restricted ventrally where differentiating cells first become evident (Fig. 2A-D). At this stage, the domains of fgfl2 and fgfl3 expression are overlapping, but not identical. Whereas fgfl2 is expressed bilaterally in the ventromedial region of spinal cord including the floor plate (Fig. 2A,B), fgfl3 is expressed in the ventrolateral horns which contain motoneurons (Fig. 2C,D). At this developmental stage, FGF expression also appears in brain outside the ependymal zones. Within the El0.5 mesencephalic region, fgfl2 and fgfl3 expression is prominent in thin bands of cells just outside the ependymal layer (Fig. 2E-G). By E12.5, fgfl2 and fgfl3 are both strongly expressed outside the ependymal zone in the spinal cord (Fig. 2H,I) and brain (data not shown) along their entire rostrocaudal and dorsoventral axes. Widespread CNS expression of both FGFs persists at E14.5 (Fig. 2L,M), with restricted regions of intense expression evident (e.g. Fig. 2J,K). fgfl3 expression declines, but is still detectable, in E16.5 CNS, while fgfl2 CNS expression is generally below detection (data not
33
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
A) mFGFI2A hFGFI2A mFGFI2B hFGFI2B mFGFI2C mFGFI3 hFGFI3 hFGFI hFGF5 hFGF9
MAAAIASSLIRQKRQARESNSDRVSASKRRSSPSKDGRSLCDRHVLGVFSKVRFCSGRKRPVRRRPEPQLKGIVTR **************************************************************************** MESK********** MESK********** .... * G R S L * S S D S P S * * * * * * * ~ * * *-*****************--***************-**********************-************** *-*****************--***************-*****--***************-************** MAEGEITTFTALTEKFNLPPGNYKKPKL MSLSFLLLLFFSHLILSAWAHG•KRLAPKGQPGPAATDRNPIGSSSRQSSSSAMSSSSASSSPAASLGSQGSGLEQSSFQWS•SGRRTGS MAPLGEVGNYFGVQDAVPFGNVPVLPVDSPVLLSDHLGQSEAGGLPRGPAVTDLDHLKGILRRRQ
mFGFI2A hFGFI2A mFGFI3 hFGFI3
LFSQQG••YFLQMH•DGTIDGTKDENSD•YTL•FNL•pV-GLRwAIQGVKASLYVA••NGEGYLYSSDVFT••CKFKESVFENYYVIYSSTLYRQQ•SGRAW ******--********************_***_******_************************************************************** ******••*•**•*•**********•*••*********************••***•***•*****•*••*****************•***•*********** ******--********************-***-******-**************************************************************
hFGFI hFGF5 hFGF9
*YCS•*-GH**RIL****•***R*R-**QHI•LQLSAESA*E--*Y*•*TETGQ*L**DT**L**G*QTPNE**L*L*•LE**H*NT*T*KKHAEKN *•CRV*IGFH**IY***K"v-N*SHEA*•L--SVLEIFAVSQ*I--*G*•**FS•KFL**SKK*K*HA*AK**DD***R*RFQ**S*NT*A*AIH*TEKT**•* *YCRT*--FH*EIF*N***•**RKDH*•FGI*•*IS*A*-**--*S*R**DSG**LG**EK*E**G*EKL*•**V*•*•FE**W*NT***•**KHVDT**RY
m F G F 12 A hFGFI2A mFGFI3 hFGFI3 hF G F 1 hFGF5 hFG F 9
FLGLNKEGPIMKG- -NRVKKTKPSSHFVPKPIEVCMYREPSLHEIGEKQGRSRKSSGTPT *************--********************************************* **
* * * * * * E* * **--***
*****
**
* * * * * * E* * **--***
*****A~.*
~.
* ***
* *********
* **
******___******
***
* *********
* **
******---*******
*V* * K * N * S C K R * - - P * T H Y G Q K A I L * L *
........ MNGGK-VVNQDST ........ *****-****** * * * ****,****** * ************
* * ********
*
* ********
*
76 76 14 14 71 71 28 90 65
173 173 168 168 .... *
122 188 162
243 243 244 244 153
L*VS
YVA* **R*KAKR*CSP***PQHI*T**L*RFKQSEQPELSFTVTVP**KNPPSPIKSKIPLSAPRKNTNSVKYRLKFRFG Y V A * * * D * T P R E * - - T * T * R H O K F T * *~ * R P V D P D K V P E L Y K D I L S Q S
268 208
B) FGF-1 100
FGF-5 40 100
FGF-9 41 45 100
FGF-12 35 35 .42 100
FGF-13 36 38 45 78 100
FGF-1 FGF-5 FGF-9 FGF-12 FGF-13
Fig. 1. FGF protein sequence alignment. (A) The predicted amino acid sequences for human and mouse FGFs are aligned, with the FGF core homology domain underlined, and with residues identical to those in murine FGF-12A indicated as '*', gaps indicated as '-', and the incomplete N-terminus of mFGF12C indicated as '...'. The sequence of murine FGF-12A (mFGF-12A) is derived from cDNA clone p2-4A (this study); human FGF-12A (hFGF-12A) (Smallwood et al., 1996); mFGF- 12B is from NCBI dbEST ID 299539 (Genbank W05845); hFGF-12B is from NCBI dbEST IDs 279066 (Genbank H19128) and 289370 (Genbank H28811) and our corrections by direct sequencing (Coulier et al., 1997); mFGF-12C from cDNA clone p2-2B (this study); mFGF-13 from cDNA clone p5-4 (this study); hFGF-13 (Smallwood et al., 1996) and correction to the N-terminal sequence based upon EMBL Z65100; hFGF-1 (Gimenez-Gallego et al., 1986), hFGF-5 (Haub and Goldfarb, 1991), and hFGF-9 (Miyamoto et al., 1993). (B) Pairwise percent amino acid sequence homology within the FGF core domain. FGF-12 and FGF-13 core domains are strongly homologous to each other and bear homology to other FGFs within the range typical for the FGF family.
shown), except in the retina, where strongfgfl2 expression persists in the retinal ganglion layer (Fig. 2N-P). fgf13 is prominently expressed from El0.5 to E16.5 throughout the peripheral nervous system, including expression in dorsal root ganglia (Fig. 2C,D,J,K and 3C), all cranial ganglia (Fig. 2E,F and 3D,E), and ganglia of the enteric nervous system (Fig. 3A,B). Fgf12 shows little or no expression in the peripheral nervous system.
2.3. Expression of FGF-12 in developing connective tissue of the skeleton
Fgfl2 expression in the cartilagenous skeleton suggests a role for this factor in the development of connective tissue. The condensations within the handplate of the fore-limb express fgf12 as early as El0.5 (Fig. 4A,B). By E12.5, fgf12 expression becomes restricted to the external portions
of phalangeal condensations (Fig. 4C,D), while the interiors of the condensations commence chondrification, as determined morphologically and by the onset of ul(II) collagen (col2al) expression (Andrikopoulos et al., 1992) (Fig. 4E,F). Similarly, fgf12 is expressed towards the exterior of the tibial condensation of the hindlimb and within a zone spanning the prospective knee joint (Fig. 4G,H), while the centers of the tibial and femoral condensations, but not the joint, express col2al (Fig. 4H,I). By E14.5, strong fg,f12 expression in the hindlimb is clearly segregated to the dense non-chondrified mesenchyme which constitutes the skeletal connective tissue (Fig. 4J,K). These new FGFs are also expressed within the trunk at presumed sites of future skeletal connective tissue. Fgf12 expression is prominent in E12.5 mesenchyme adjacent to vertebral neural arch chondrifications (Fig. 4L,M) and in mesenchyme separating vertebrae from the base of rib shafts
34
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
(Fig. 4N-Q). Skeletal expression offgfl2 is not detectable by embryonic day 16, and fgf13 is not coexpressed at any sites of developing connective tissue.
2.4. FGF-12 and FGF13 expression in developing heart Fgf12 and fgf13 are expressed in the myocardium of the developing heart from El0.5 through E14.5. Sagittal sections through the heart of El0.5 and El2.5 embryos shows fgfl2 expression restricted to the myocardium of the atrial chamber (Fig. 5A,B,E). By contrast, fgf13 is expressed in both atrial and ventricular myocardia, but is not expressed in endocardial cushions (Fig. 5C,D). 2.5. Other sites of FGF-12 and FGF-13 embryonic expression 2.5.1. Axial skeleton When primitive sclerotomal tissue becomes rostrocaudally organized into alternating bands of looser and denser mesenchyme, fgf12 is expressed in the bands of loose mesenchyme. Upon further reorganization of the sclerotomal tissue into chondrogenic vertebral body masses and intervertebral tissue, fgf12 expression is downregulated. This is illustrated in the caudal region at embryonic day 14.5 (Fig. 6A,B), where fgf12 is strongly expressed in the bands of looser sclerotomal mesenchyme in the tail, but expression has diminished in more rostral (sacral) segments which have undergone reorganization. 2.5.2. Craniofacial mesenchyme Fgf12 is expressed in regions of craniofacial mesenchyme between El0.5 and E14.5. As one example, basiooccipital mesenchyme just beneath the brain bears segments of fgf12 expression at El0.5 (Fig. 2G) and E12.5 (Fig. 6E,F).
Fig. 2. Expression of murine fgfl2 and fgfl3 RNAs in embryonic central nervous system. Deparaffinized embryo sections were hybridized with 35Slabelled antisense riboprobes for murinefgcf12 orfgfl3, exposed to photographic emulsion, stained, and viewed under bright-field (A,C,E,H,J,N) or dark-field (B.D,F,G+I,K-M,O,P) illumination. (A,B)fgff12, El0.5, spinal cord, transverse section, expression in floor plate and ventromedial spinal cord; (C,D)fgf13, El0.5, spinal cord, transverse, expression in ventral horns and dorsal root ganglia; (E,F)fgf13, El0.5, head, longitudinal. expression in trigeminal ganglia and midbrain outside ependymal layer (arrowheads); (G)fgoC12. El0.5, head, longitudinal, expression outside ependymal layer (arrowheads) and in basiooccipital mesenchyme (arrows); (H,I) fgfl2, E12.5, spinal cord, transverse, expression in floor plate and outside ependymal layer; (J,K)fgcf13, E 14.5, spinal cord, transverse, broad expression with regionally strong expression; (L)fgf12, E14.5, whole embryo, sagittal, expression throughout brain and spinal cord; (M) FGF13, E14.5, whole embryo, sagittal, expression throughout brain and spinal cord; (N,O)fgf12, E16.5, eye, sagittal, strong expression in retinal ganglion layer and weak expression in lens; (P)fgf13, E16.5, eye, sagittal, no detectable expression. Spinal cord (sc), ependymal layer (el, floor plate (f), ventral horn (vh/, dorsal root ganglion (d), midbrain (rob), trigeminal ganglion (t), oropharynx (p). vertebral body (v), neuroepithelium (n), retinal ganglion layer (g), lens (le).
......
:
+
.
:~I+
....
•
+
.
H. Hartunget al. /Mechanismsof Development64 (1997)31-39
35
Fig. 3. Fgf13RNA in E14.5 peripheral nervous system. Bright field (A,C,D) and dark field (B,E) offgfl3 hybridization to sections of E14.5 embryo. (A,B) Intestine, cross-section,expressionin enteric ganglia situated between inner circular and outer longitudinal muscle layers; (C) strong expression in dorsal root ganglia; (D,E) head, parasagittal section, strong expression in fifth, seventh, eighth, and ninth cranial ganglia. Luminal epithelium (e), mesenchymallayer (m), inner muscular layer (i), outer muscular layer (o), dorsal root ganglia (d), vertebral body (v), canals of inner ear (c), cranial ganglia (5,7,8,9), brain (b).
2.5.3. Nasal epithelium Fgf12 is expressed in nasal olfactory epithelium at E12.5 (Fig. 6C,D) and E14.5 (data not shown).
2.5.4. Kidney Fgf12 andfgf13 are expressed in non-overlapping regions of the developing kidney. Whereas fgf12 expression is restricted to glomeruli (Fig. 6E,F),fgfl3 is expressed in patches of loose mesenchyme surrounding the collecting duct and near glomeruli (Fig. 6G,H).
16 tissues analyzed. By contrast, fgf12 R N A species were seen in heart, brain, prostate, ovaries, and intestine, but not in other tissues. The 6.0, 2.6, and 1.2 kb species offgfl2 RNA seen in varying relative abundance may result from alternative pre-mRNA splicing, as described earlier in the analysis of fgf12 c D N A clones. These findings stand in sharp contrast to other FGF genes, whose expression in adult tissues is highly restricted or completely absent (Dickson et al., 1989; Curatola and Basilico, 1990; de Lapeyriere et al., 1990).
2.6. Expression in adult tissues 3. Discussion The expression offgfl2 andfgfl3 genes in adult human tissues was analyzed by Northern blot hybridization with 32p-labelled human FGF cDNA fragments. As shown in Fig. 7, a 2.1 kb fgf13 RNA transcript was detectable in all
Fgfl2 andfgfl3 genes are dynamically expressed in midgestation mouse embryos at sites different from those described for other previously characterized FGFs. While
36
H. Hartung et al. / Mechanisms of Development 64 (1997) 31-39
Fig. 4. Fgfl2 RNA in developing skeletal connective tissue. Hybridizations with fgfl2 or col2al viewed in bright (A,C,E,H,J,L,N,P) or dark field (B,D,F,G,I,K,M,O,Q). (A,B)fgfl2, El0.5, handplate, cross-section, expression condensations; (C,D)fgf12, E12.5, handplate, cross-section, expression in exterior of phalangeal condensations; (E,F) col2al, section adjacent to (C,D), expression in interior of condensations: (G)fgf12, E12.5, hindlimb, oblique, expression in exterior of tibial condensation and in knee joint: (H,I) col2al, section adjacent to (G), expression within tibial and femural condensations; (J,K) fgf12, E14.5, hindlimb, sagittal, expression in dense connective tissue mesenchyme and virtual exclusion from cartilagenous condensations (*); (L,M)fgf12, E12.5, cervical region, expression in mesenchyme near vertebral neural arches; (N,O) parasagittal and (P,Q) longitudinal sections, E12.5,fgf12, expression in mesenchyme between chondrifications of vertebral bodies and bases of rib shafts. Forelimb (ft), hindlimb (hi), prospective tibia (ti), femur (fe), knee joint (j), vertebral neural arch (n), vertebral body (v), rib shaft (r), spinal cord (sc).
H. Hartung et aL /Mechanisms of Development 64 (1997) 31-39
37
fgf3 and fgf8 are transiently expressed within restricted domains of brain neuroepithelium (Wilkinson et al., 1989; Crossley and Martin, 1995), fgfl2 and fgfl3 are, so far, unique among FGFs in their prominent and widespread expression throughout the embryonic CNS and PNS. Similarly, expression offgfl2 andfgfl3 in developing heart and connective tissue is, so far, unique among FGFs. These observations suggest several novel embryonic functions for these newly identified FGFs. The use of gene targetting in mice to disruptfgfl2 orfgfl3 expression and of transgenic methods to promote regional ectopicfg(12 orfgfl3 expression should reveal certain embryonic requirements for these factors. Several enigmatic properties of FGF-12 and FGF-13, along with closely related FGF-11 (FHF-3) and FHF-4, complicate speculation regarding their biological and biochemical functions. These FGFs form a distinct sublineage of FGFs characterized by strong homology within the FGF core domain along with substantial homology outside of the core. While recombinant FGF-12 and FGF-13 proteins share with other FGFs the ability to bind avidly to heparin-agarose, FGF-12 and FGF-13 cannot activate many of the known FGF receptor tyrosine kinases (H. Hartung and M. Goldfarb, unpublished data). Additionally, FGF-12 and FGF-13 fail to mimic the ability of other FGFs to induce mesoderm in animal caps excised from blastula-stage frog embryos (A. Suzuki, H. Hartung, M. Goldfarb and A. Hemmati-Brivanlou, unpublished data). Furthermore, no members of the new FGF subfamily bear N-terminal signal sequences for secretion, and FGF-12/ FHF-1 protein has been shown to accumulate intracellularly in vivo (Smallwood et al., 1996). It is possible that these new FGFs activate novel cell surface receptors or interact non-productively (antagonistically) with known FGF receptors. Alternatively, the FGF core homology domain may represent a functionally versatile structural module utilized by FGF-12 and FGF-13 for novel molecular interactions within the cell.
4. Materials and methods
4.1. Cloning and sequencing murine FGF-12 and FGF-13 cDNAs
Fig. 5. Fg~12 and fgfl3 RNAs in embryonic heart. Detection of fgfl2 (A,B,E) and fgfl3 (C,D) RNAs in parasagittal sections of El0.5 (A,B) or E12.5 (C-E) heart viewed under bright field (A,C) or dark field (B,D,E). Atrial myocardiurn (a), ventricular myocardium (v), endocardial cushions (e), and liver (li).
The complete coding sequence of a human fgfl2 cDNA clone (Coulier et al., 1997) (American Type Culture Collection, IMAGE Consortium) was amplified by polymerase chain reaction (PCR), 32p-labelled with random hexamer primers and DNA polymerase I Klenow fragment, and used to screen a whole mouse E17.5 cDNA lamda ZAP library (Stratagene) by filter transfer plaque hybridization. Purified positive clones were converted to plasmids by in vivo excision and their identities determined by PCR with FGF-specific primers followed by dideoxynucleotide sequencing. The coding sequences of murine fgfl3 cDNA
H. Hartung et al./ Mechanisms of Development 64 (1997) 31-39
38
i
C(;:;
• ::•2
Fig. 6. Other sites offgfl2 andfgf13 embryonic expression. Bright field (A,C,E,G,I) and dark field (B,D,F,H,J) illumination. (A,B)fgfl2, E14.5, caudal region, mid-sagittal, segmental expression in the loose mesenchyme bands (brackets) of tail sclerotomes but diminishing in segments undergoing vertebral differentiation; (C,D)fgf12, E12.5, snout, parasagittal, expression in nasal olfactory epithelium; (E,F)fgfl2, E12.5, head, longitudinal, expression in basiooccipital region above tongue; (G,H)fgfl2 and (I,J)fgfl3, E14.5, kidney, parasagittai, withfgf12 RNA in glomeruli andfgfl3 RNA in mesenchyme. Sacral vertebral body (v), brain vesicle (bv), nasal epithelium (n), snout (sn), tongue (t), basiooccipital region (bo), kidney collecting duct (d), glomeruli (g).
h
b
pl lu Ii sm k pa
~
$p th IX te o si c phi
kb
$p th pr te o si ¢ phi
m mmm
FGF12
e h
1
b
131 lu Ii sm k pa . . . . . .
~
O.s..-~:~:~
: i ii~ ~
~
FGF13
Fig. 7. Northern blot analysis offgfl2 andfgfl3 expression in adult human tissues. Human polyA + RNA from different tissues were electrophoresed through parafbrmaldehyde agarose gels, transferred to nylon membranes, and hybridized with 32p-labelled fragments of humanfgfl2 orfg~13 cDNAs. Washed filters were autoradiographed for 3-7 days. Heart (h), brain (b), placenta (pl), lung (lu), liver (li) skeletal muscle (sm), kidney (k), pancreas (p), spleen (sp), thymus (th), prostate (pr), testis (te), ovary (o), small intestine (si), colon (c), and peripheral blood lymphocyte (PBL). The sizes of molecular weight markers are indicated.
1t. Hartung et al. / Mechanisms of Development 64 (1997) 31-39 clone p5-4, mfgfl2A c D N A clone 2-4A, and the alternative N-terminal coding sequence of mfgf12C c D N A clone 2-2B have been deposited in Genbank.
4.2. In situ hybridization Segments ofmurinefgH2,fgfl3, and col2al cDNAs were used as templates for antisense-strand 35S-labelled riboprobe synthesis with T7 R N A polymerase. Templates were as follows: mfgf12, 1.0 kb whole c D N A insert of plasmid p2-2B; mfgf13, 1.6 kb fragment of p5-4 c D N A insert, containing 200 bp C-terminal coding sequence and 1.4 kb 3' untranslated sequence; col2al, 450 bp fragment of 3' untranslated sequence (Andrikopoulos et al., 1992). Hybridization of probes to deparaffinized sections of mouse embryos was performed by the method of Wilkinson (Wilkinson and Nieto, 1993), with modification of the hybridization solution by inclusion of 20 m M sodium phosphate (pH 7.5). After hybridization, slides were dipped in undiluted Kodak NTB2 emulsion, exposed 10-21 days (except 5 days for col2al probe), developed and stained with hematoxylin/eosinY. Bright- and dark-field images viewed on Nikon Microphot and S M Z - U microscopes were captured with a Kodak DCS410 digital camera.
4.3. Northern blot hybridization Human multiple tissue Northern blots (MTN and MTNII, Clontech Laboratories) were hybridized and washed under manufacturer's conditions using fragments o f F G F cDNAs 32p-labelled by random hexamer priming as probes. Probes were as follows: fgf12, 0.9 kb EcoRUNotI insert f r o r n plasmid yr45d03 (Genbank ID H62672); fgfl3, 0.75 kb PstI fragment from plasmid yg96e06 (Genbank ID R59031). Washed blots were exposed to Kodak X A R film for 3 - 7 days.
Acknowledgements W e thank Sam Bouyain for excellent assistance with D N A sequencing. M.G. is a research fellow of the Lucille P. Markey Charitable Trust.
39
References Amaya, E., Musci, T.J. and Kirschner, M.W. (1991) Cell 66, 257-270. Andrikopoulos,K., Suzuki, H.R., Solursh, M. and RamireL F. (1992) Dev. Dyn. 195, 113-120. Coulier, F., Pontarotti, P., Roubin, R., Hartung, H., Goldfarb M. and Birnbaum, D. (1997) J. Mol. Evol. 44, 43-56. Cox, W.G. and Hemmati-Brivanlou, A. (1995) Development 121, 43494358. Crossley, P.H. and Martin, G.R. (1995) Development 121, 439-451. Curatola, A.M. and Basilico, C. (1990) Mol. Cell. Biol. 10, 2475-2484. de Lapeyriere, O., Rosnet, O., Benharroch, D., Raybaud, F., Marchetto, S., Planche, J., Galland, F., Mattei, M.-G., Copeland, N.G., Jenkins, N.A., Coulier, F. and Birnbaum, D. (1990) Oncogene 5, 823-831. Deng, C.-X., Wynshaw-Boris, A., Shen, M.M., Daugherty, C., Omitz, D.M. and Leder, P. (1994) Genes Dev. 8, 3045-3057. Dickson, C., Deed, R., Dixon, M. and Peters, G. (1989) Prog. Growth Factor Res. 1. 123-132. Faham, S., Hileman, R.E., Fromm, J.R., Linhardt, R.J. and Rees, D.C. (1996) Science 271, 1116-1120. Feldman, B., Pouymeirou, W., Papaioannou, V.E., DeChiara, T. and Goldfarb, M. (1995) Science 267, 246-249. Gimenez-Gallego, G., Conn, G., Hatcher, V.B. and Thomas, K.A. (1986) Biochem. Biophys. Res. Commun. 138, 611-617. Goldfarb, M. (1996) Cytokine Growth Factor Rev. 7, 311-325. Haub, O. and Goldfarb, M. (1991) Development 112, 397-406. Kengaku, M. and Okamoto, H. (1995) Development 121, 3121-3130. Lamb, T.M. and Harland, R.M. (1995) Development 121. 3627-3636. Miyamoto, M., Naruo, K.-I., Seko, C., Matsumoto, S., Kondo, T. and Kurokawa, T. (1993) Mol. Cell. Biol. 13, 4251-4259. Rapraeger, A.C., Krufka, A. and Olwin, B.B. (1991) Science 252, 17051708. Shing, Y., Folkman, J., Sullivan, R., Butterfield, C., Murray, J. and Klagsbrun, M. (1984) Science 223, 1296-1299. Smallwood, P.M.. Munoz-Sanjuan, I., Tong, P., Macke, J.P., Hendry, S.H.C., Gilbert, D.J., Copeland, N.G., Jenkins, N.A. and Nathans, J. (1996) Proc. Natl. Acad. Sci. USA 93, 9850-9857. Verdier, A.-S., Mattei, M.-G., Lovec, H., Hartung, H., Goldfarb, M., Bimbaum, D. and Coulier, F. (1997) Genomics 40, 151-154. Wilkinson, D.G., Bhatt, S. and McMahon, A.P. (1989) Development 105, 131-136. Wilkinson. D.G. and Nieto, M.A. (1993) Methods Enzymol. 225, 361373. Yamaguchi. T.P., Harpal, K., Henkemeyer, M. and Rossant, J. (1994) Genes Dev. 8, 3032-3044. Yayon, A., Klagsbrun, M.. Esko, J.D., Leder, P. and Omitz, D.M. (1991) Cell 64, 841-848. Zhu, X., Komiya, H., Chirino, A., Faham, S., Fox, G.M., Arakawa, T., Hsu, B.T. and Rees, D.C. (1991) Science 251.90-93.