- Email: [email protected]
The mouse Fgfrl1 gene coding for a novel FGF receptor-like protein1
Biochimica et Biophysica Acta 1520 (2001) 247^250
www.bba-direct.com
Short sequence-paper
The mouse Fgfrl1 gene coding for a novel FGF receptor-like protein1 Markus Wiedemann, Beat Trueb * M.E. Mu«ller-Institute, University of Bern, P.O. Box 30, CH-3010 Bern, Switzerland Received 11 April 2001; received in revised form 18 June 2001; accepted 20 June 2001
Abstract The mouse Fgfrl1 gene codes for a novel cell surface protein that is closely related to the family of the FGF receptors. It contains three extracellular Ig C2 loops and an acidic box, which share 29^33% sequence identity (48^50% similarity) with FGF receptors 1^4. The intracellular portion of the novel protein, however, lacks a tyrosine kinase domain required for signal transduction by transphosphorylation. The gene for Fgfrl1 comprises six exons and is located on mouse chromosome 5 in close proximity to the Idua gene for L-iduronidase. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Bone development; Cartilage ; Endochondral ossi¢cation ; Fibroblast growth factor; FGF receptor ; Gene structure; Mouse
The ¢broblast growth factors (FGFs) and their corresponding receptors (FGFRs) control the proliferation, differentiation and migration of most mesenchymal and neuroectodermal cells [1^3]. At present, 23 di¡erent FGF ligands which activate four di¡erent FGF receptors (FGFR1-FGFR4) are known. The activity of the FGFs is dramatically increased by heparin or heparan sulfate proteoglycans. A complex consisting of two FGFRs, one heparan sulfate glycosaminoglycan chain and two FGF ligands seems to represent the physiologically active form [4]. We have been interested for some time in the structure and regulation of cartilage proteins. To search for novel regulatory proteins we have prepared a subtracted cDNA library enriched in cartilage-speci¢c genes [5]. One of the clones from this library encoded a receptor-like protein, which might further contribute to the complexity of the FGF signaling system [6]. The novel protein, termed FGFR-like protein 1 (FGFRL1), contains three extracellular Ig-like domains as well as a transmembrane segment that share signi¢cant similarity with the extracellular and transmembrane domains of all FGFRs. The novel protein, however, lacks an intracellular tyrosine kinase domain and consequently will not be able to signal by transphosphor* Corresponding author. Fax: +41-31-632-4999. E-mail address : [email protected] (B. Trueb). 1 The nucleotide sequence data reported in this paper have been submitted to the DDBJ, EMBL, GenBank databases and appear under accession numbers AJ293947 and AJ308490.
ylation. Based on the substantial similarity with members of the FGFR family, we have speculated that it might interact with a true FGFR, thereby modulating or inhibiting its activity [6]. It is also conceivable that the novel protein might bind FGF ligand and sequester it from the plasma membrane. Recently, the same protein has independently been discovered by Kim et al. and termed FGFR5 [7]. Animal models with a targeted disruption in a particular gene have largely contributed to our current understanding of the function of extracellular matrix proteins [8] during limb development. To enable the generation of such a knock-out mouse we have set out to clone the murine Fgfrl1 gene. Three mouse EST clones that showed high similarity to the human FGFRL1 cDNA sequence were selected from the GenEMBL databank and ordered from the Resource Center of the German Human Genome Project (IMAGp998F012103, IMAGp998P06933, IMAGp998C22756). Sequencing studies demonstrated that the three clones spanned a total of 2300 bp encoding the entire mouse Fgfrl1 protein (accession number AJ293947). The cDNA sequence contained 94 bp of 5P UTR, 1587 bp of open reading frame, 590 bp of 3P UTR as well as a short poly(A) tail of 29 nucleotides. The open reading frame could be translated into a protein of 529 amino acids with a molecular mass of 57 kDa and an isoelectric point of 10.4. Compared to human FGFRL1 [6], the mouse sequence showed 77% identity at the nucleotide and 88% identity at the protein level (Fig. 1). Predictions with the
0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 2 6 7 - 6
BBAEXP 91567 4-9-01
Cyaan Magenta Geel Zwart
248
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
Fig. 1. Alignment of the amino acid sequences from mouse and human FGFRL1 and mouse FGFR3. Identical residues are boxed. The three Ig C2 loops are indicated by brackets, the transmembrane domain is shown by a dotted box. The signal peptide cleavage site is marked by an arrow.
computer program SMART [9] demonstrated that the mouse Fgfrl1 protein contained all the typical features of the extracellular domains of FGFRs. It was composed of
Fig. 2. Expression of the mouse Fgfrl1 mRNA at di¡erent developmental stages. A Northern blot containing poly(A) RNA (2 Wg/lane) from mouse embryos of four developmental stages (7, 11, 15, and 17 days, Clontech Laboratories) was hybridized successively with radiolabeled probes for Fgfrl1, Fgfr1, and Gapdh as indicated. The migration position of RNA standards (kb) is shown in the left margin.
BBAEXP 91567 4-9-01
a hydrophobic signal peptide that could be cleaved after Ala(20) with a score of 14.6 [10], of three extracellular Ig C2 loops, of a hydrophilic box inserted between the ¢rst and second Ig C2 loop, and of a hydrophobic segment that is likely to form a transmembrane domain. In con-
Fig. 3. Exon/intron structure of the mouse Fgfrl1 gene. Exons 1^6 of the mouse Fgfrl1 gene and exon 14 of the mouse Idua gene are indicated. The frequency of the dinucleotide sequence CpG in a window of 300 nucleotides is included at the top.
Cyaan Magenta Geel Zwart
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
249
Table 1 Exon/intron structure of the mouse Fgfr11 gene Exon
Position on cDNA
Splice junctions
Domain encoded
Acceptor site 1
1^161
2
162^434
actcttgcag
3
435^515
ttttctacag
4
516^800
tgccccacag
5
801^1154
cttggcttag
6
1155^2270
tctcttgcag
Intron (bp)
Donor site ... GAGGCGCGAG AlaAlaArgG ... ATCATCATGG IleIleMetA ... CAGCAGTGGG GlnGlnTrpA ... GATGTAATCC AspValIleG ... GTATTACCAG ValLeuProA ... GTTAAACAAG
GACCCCCAAG lyProProAr ATGATATTAG spAspIleSe CACGGCCTCG laArgProAr AGCGGACTCG lnArgThrAr ACCCCAAACC spProLysPr
trast to all FGFRs, however, it lacked the intracellular tyrosine kinase domain, but instead possessed 134 unrelated residues at its C-terminus. The extracellular portion of the novel protein (residues 40^352) shared 29^33% sequence identity (48^50% sequence similarity) with mouse Fgfr1^Fgfr4 (Fig. 1). The intracellular portion, however, showed less than 20% sequence identity with any other protein of the Swissprot databank. The expression of the novel gene was examined on a Northern blot containing poly(A) RNA from total mouse embryos of four developmental stages (Fig. 2). A radiolabeled cDNA probe for Fgfrl1 hybridized speci¢cally to a mRNA of V2600 nucleotides consistent with the size of the mouse cDNA sequence assuming a poly(A) tail of 300 nucleotides. This mRNA was barely detectable at embryonic day 7, but its signal became clearly visible at day 11 and increased in intensity until day 17. The developmental expression of Fgfrl1 di¡ered signi¢cantly from that of Fgfr1 which was highly expressed at embryonic day 7, but barely expressed at later stages. Hybridization of the same Northern blot with probes speci¢c for mouse
gtgagt
5P UTR, signal
gtcagt
Ig C2
gtaagt
Acidic Box
gtgagt
Ig C2
gtgtgt
Ig C2
(A)n
TM, intracellular, 3P UTR
8166 1108 107 79 243
Fgfr2, Fgfr3, and Fgfr4 did not yield any distinct signals, suggesting very low expression of these genes in total mouse embryos. The mouse cDNA clones were used to screen a genomic DNA library prepared from mouse liver in the vector EMBL3 (Clontech Laboratories ML1030j). After two rounds of screening, a total of nine genomic clones were obtained which together spanned the entire mouse Fgfrl1 gene. Restriction enzyme mapping and direct sequencing of 24 kb of genomic DNA (accession number AJ308490) established the complete exon/intron structure of the mouse gene (Fig. 3). This gene was found to consist of six exons and ¢ve introns. The ¢rst exon which was separated from the rest of the gene by a large intron of 8,2 kb contained the 5P UTR and the sequence for the signal peptide (Table 1). The remaining exons were clustered within 4 kb. Each of these exons coded for a complete, functional domain. Three exons coded for the three IgC2 domains, one coded for the acidic box, and the last exon coded for the transmembrane domain and the short Cterminal tail (Table 1). The ¢rst exon was situated within
Fig. 4. Localization of the Fgfrl1 gene on mouse chromosomes by FISH. The FISH signal is shown at the left, the DAPI banding pattern in the middle. The position of the gene was assigned to region 5E3-F by superimposing the FISH signal and the DAPI banding pattern from ten independent determinations as shown on the right.
BBAEXP 91567 4-9-01
Cyaan Magenta Geel Zwart
250
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
a typical CpG island where the dinucleotide sequence CG occurred with high frequency relative to the rest of the genome (Fig. 3). CpG islands are characteristic features of the promoters for housekeeping genes and many proto-oncogenes [11]. When the upstream region of our genomic sequence was carefully inspected, part of the 3P end from the Idua gene was identi¢ed. Hence, the Fgfrl1 gene is separated from the Idua gene by just 10 kb and both genes point into the same direction. The Idua gene is known to encode L-iduronidase and to consist of 14 exons [12,13]. Mutations in this gene are the principal cause of mucopolysaccharidosis type I [14]. Restriction mapping indicated that our genomic clones did not comprise the entire Idua gene, but only exons 11^14. The intervening sequence between the Idua and the Fgfrl1 gene harbored several repetitive DNA elements. There was one copy of the IAPLTR1 MM element (position 2278^2609), one copy of a novel element (position 4899^5045) and two copies of the MMB2 element (positions 6377^6568, 6961^7148). The signi¢cance of these repetitive elements is not known. The location of the novel gene (approved gene symbol Fgfrl1) was mapped on mouse chromosomes by the FISH technique (Fig. 4). A biotinylated restriction fragment of 5.5 kb encompassing exons 2^6 (positions 18 686^24 173) was hybridized to mouse chromosomes prepared from spleen lymphocytes [15,16]. This probe bound speci¢cally to mouse chromosome 5 as demonstrated by staining with £uorescently labeled avidin. Among 100 metaphase spreads analyzed, 87 contained at least two hybridization signals per pair of chromosomes. The exact location as determined by superimposing the FISH signal and the DAPI banding pattern was found to be region E3-F (Fig. 4). This result is consistent with the location of the human FGFRL1 gene that has been mapped to human chromosome 4 band 4p16 [6]. Region 4p16^4q21 is syntenic with region 18^57 cM of mouse chromosome 5, constituting homology group 38 (see Human-Mouse Homology Map at www.ncbi.nlm.nih/Homology). The mouse Idua gene which according to our sequencing results is situated just 10 kb upstream from the Fgfrl1 gene has been placed to chromosome 5 position 57.0 cM by previous genetic analyses. Our mapping data thus allow the correlation of the cytogenetic region 5 E3-F with the genetic region 5 57.0 cM. This region must have undergone substantial rearrangement during evolution of mice and men, because the three human genes FGFRL1, IDUA and FGFR3 are clustered in chromosomal band 4p16, whereas the mouse Fgfrl1 and Idua genes are situated at
BBAEXP 91567 4-9-01
position 5 57 cM, but Fgfr3 is found far apart at position 5 20 cM. Our localization studies might aid in the ¢nal e¡orts to establish a comprehensive map of the mouse genome and to unravel its complete DNA sequence. Our data should also permit the generation of a transgenic animal with a targeted disruption of the Fgfrl1 gene. This study was supported by the Swiss National Science Foundation (Grant 31-61296.00).
References [1] W.L. McKeehan, F. Wang, M. Kan, The heparan sulfate-¢broblast growth factor family : diversity of structure and function, Prog. Nucleic Acid Res. Mol. Biol. 59 (1998) 135^176. [2] G. Szebenyi, J.F. Fallon, Fibroblast growth factors as multifunctional signaling factors, Int. Rev. Cytol. 185 (1999) 45^106. [3] D.M. Ornitz, N. Itoh, Fibroblast growth factors, Genome Biol. 2 (2001) 3005.1^3005.12. [4] L. Pellegrini, D.F. Burke, F. von Delft, B. Mulloy, T.L. Blundell, Crystal structure of ¢broblast growth factor receptor ectodomain bound to ligand and heparin, Nature 407 (2000) 1029^1034. [5] D. Belluoccio, B. Trueb, Matrilin-3 from chicken cartilage, FEBS Lett. 415 (1997) 212^216. [6] M. Wiedemann, B. Trueb, Characterization of a novel protein (FGFRL1) from human cartilage related to FGF receptors, Genomics 69 (2000) 275^279. [7] I. Kim, S.-O. Moon, K.-H. Yu, U.-H. Kim, G.Y. Koh, A novel ¢broblast growth factor receptor-5 preferentially expressed in the pancreas, Biochim. Biophys. Acta 1518 (2001) 152^156. [8] E. Gustafsson, R. Fa«ssler, Insights into extracellular matrix functions from mutant mouse models, Exp. Cell Res. 261 (2000) 52^68. [9] J. Schultz, R.R. Copley, T. Doerks, C.P. Ponting, P. Bork, SMART : a web-based tool for the study of genetically mobile domains, Nucleic Acids Res. 28 (2000) 231^234. [10] G. Von Heijne, A new method for predicting signal sequence cleavage sites, Nucleic Acids Res. 14 (1986) 4683^4690. [11] S.H. Cross, A.P. Bird, CpG islands and genes, Curr. Opin. Genet. Dev. 5 (1995) 309^314. [12] H.S. Scott, X.-H. Guo, J.J. Hopwood, C.P. Morris, Structure and sequence of the human alpha-L-iduronidase gene, Genomics 13 (1992) 1311^1313. [13] L.A. Clarke, J. Nasir, H. Zhang, H. McDonald, D.A. Applegarth, M.R. Hayden, J. Toone, Murine alpha-L-iduronidase: cDNA isolation and expression, Genomics 24 (1994) 311^316. [14] H.S. Scott, S. Bunge, A. Gal, L.A. Clarke, C.P. Morris, J.J. Hopwood, Molecular genetics of mucopolysaccharidosis type I: diagnostic, clinical, and biological implications, Hum. Mutat. 6 (1995) 288^ 302. [15] H.H.Q. Heng, J. Squire, L.-C. Tsui, High-resolution mapping of mammalian genes by in situ hybridization to free chromatin, Proc. Natl. Acad. Sci. USA 89 (1992) 9509^9513. [16] M. Imhof, B. Trueb, Comparative cytogenetic mapping of COL14A1, the gene for human and mouse collagen XIV, Cytogenet. Cell Genet. 84 (1999) 217^219.
Cyaan Magenta Geel Zwart
www.bba-direct.com
Short sequence-paper
The mouse Fgfrl1 gene coding for a novel FGF receptor-like protein1 Markus Wiedemann, Beat Trueb * M.E. Mu«ller-Institute, University of Bern, P.O. Box 30, CH-3010 Bern, Switzerland Received 11 April 2001; received in revised form 18 June 2001; accepted 20 June 2001
Abstract The mouse Fgfrl1 gene codes for a novel cell surface protein that is closely related to the family of the FGF receptors. It contains three extracellular Ig C2 loops and an acidic box, which share 29^33% sequence identity (48^50% similarity) with FGF receptors 1^4. The intracellular portion of the novel protein, however, lacks a tyrosine kinase domain required for signal transduction by transphosphorylation. The gene for Fgfrl1 comprises six exons and is located on mouse chromosome 5 in close proximity to the Idua gene for L-iduronidase. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Bone development; Cartilage ; Endochondral ossi¢cation ; Fibroblast growth factor; FGF receptor ; Gene structure; Mouse
The ¢broblast growth factors (FGFs) and their corresponding receptors (FGFRs) control the proliferation, differentiation and migration of most mesenchymal and neuroectodermal cells [1^3]. At present, 23 di¡erent FGF ligands which activate four di¡erent FGF receptors (FGFR1-FGFR4) are known. The activity of the FGFs is dramatically increased by heparin or heparan sulfate proteoglycans. A complex consisting of two FGFRs, one heparan sulfate glycosaminoglycan chain and two FGF ligands seems to represent the physiologically active form [4]. We have been interested for some time in the structure and regulation of cartilage proteins. To search for novel regulatory proteins we have prepared a subtracted cDNA library enriched in cartilage-speci¢c genes [5]. One of the clones from this library encoded a receptor-like protein, which might further contribute to the complexity of the FGF signaling system [6]. The novel protein, termed FGFR-like protein 1 (FGFRL1), contains three extracellular Ig-like domains as well as a transmembrane segment that share signi¢cant similarity with the extracellular and transmembrane domains of all FGFRs. The novel protein, however, lacks an intracellular tyrosine kinase domain and consequently will not be able to signal by transphosphor* Corresponding author. Fax: +41-31-632-4999. E-mail address : [email protected] (B. Trueb). 1 The nucleotide sequence data reported in this paper have been submitted to the DDBJ, EMBL, GenBank databases and appear under accession numbers AJ293947 and AJ308490.
ylation. Based on the substantial similarity with members of the FGFR family, we have speculated that it might interact with a true FGFR, thereby modulating or inhibiting its activity [6]. It is also conceivable that the novel protein might bind FGF ligand and sequester it from the plasma membrane. Recently, the same protein has independently been discovered by Kim et al. and termed FGFR5 [7]. Animal models with a targeted disruption in a particular gene have largely contributed to our current understanding of the function of extracellular matrix proteins [8] during limb development. To enable the generation of such a knock-out mouse we have set out to clone the murine Fgfrl1 gene. Three mouse EST clones that showed high similarity to the human FGFRL1 cDNA sequence were selected from the GenEMBL databank and ordered from the Resource Center of the German Human Genome Project (IMAGp998F012103, IMAGp998P06933, IMAGp998C22756). Sequencing studies demonstrated that the three clones spanned a total of 2300 bp encoding the entire mouse Fgfrl1 protein (accession number AJ293947). The cDNA sequence contained 94 bp of 5P UTR, 1587 bp of open reading frame, 590 bp of 3P UTR as well as a short poly(A) tail of 29 nucleotides. The open reading frame could be translated into a protein of 529 amino acids with a molecular mass of 57 kDa and an isoelectric point of 10.4. Compared to human FGFRL1 [6], the mouse sequence showed 77% identity at the nucleotide and 88% identity at the protein level (Fig. 1). Predictions with the
0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 2 6 7 - 6
BBAEXP 91567 4-9-01
Cyaan Magenta Geel Zwart
248
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
Fig. 1. Alignment of the amino acid sequences from mouse and human FGFRL1 and mouse FGFR3. Identical residues are boxed. The three Ig C2 loops are indicated by brackets, the transmembrane domain is shown by a dotted box. The signal peptide cleavage site is marked by an arrow.
computer program SMART [9] demonstrated that the mouse Fgfrl1 protein contained all the typical features of the extracellular domains of FGFRs. It was composed of
Fig. 2. Expression of the mouse Fgfrl1 mRNA at di¡erent developmental stages. A Northern blot containing poly(A) RNA (2 Wg/lane) from mouse embryos of four developmental stages (7, 11, 15, and 17 days, Clontech Laboratories) was hybridized successively with radiolabeled probes for Fgfrl1, Fgfr1, and Gapdh as indicated. The migration position of RNA standards (kb) is shown in the left margin.
BBAEXP 91567 4-9-01
a hydrophobic signal peptide that could be cleaved after Ala(20) with a score of 14.6 [10], of three extracellular Ig C2 loops, of a hydrophilic box inserted between the ¢rst and second Ig C2 loop, and of a hydrophobic segment that is likely to form a transmembrane domain. In con-
Fig. 3. Exon/intron structure of the mouse Fgfrl1 gene. Exons 1^6 of the mouse Fgfrl1 gene and exon 14 of the mouse Idua gene are indicated. The frequency of the dinucleotide sequence CpG in a window of 300 nucleotides is included at the top.
Cyaan Magenta Geel Zwart
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
249
Table 1 Exon/intron structure of the mouse Fgfr11 gene Exon
Position on cDNA
Splice junctions
Domain encoded
Acceptor site 1
1^161
2
162^434
actcttgcag
3
435^515
ttttctacag
4
516^800
tgccccacag
5
801^1154
cttggcttag
6
1155^2270
tctcttgcag
Intron (bp)
Donor site ... GAGGCGCGAG AlaAlaArgG ... ATCATCATGG IleIleMetA ... CAGCAGTGGG GlnGlnTrpA ... GATGTAATCC AspValIleG ... GTATTACCAG ValLeuProA ... GTTAAACAAG
GACCCCCAAG lyProProAr ATGATATTAG spAspIleSe CACGGCCTCG laArgProAr AGCGGACTCG lnArgThrAr ACCCCAAACC spProLysPr
trast to all FGFRs, however, it lacked the intracellular tyrosine kinase domain, but instead possessed 134 unrelated residues at its C-terminus. The extracellular portion of the novel protein (residues 40^352) shared 29^33% sequence identity (48^50% sequence similarity) with mouse Fgfr1^Fgfr4 (Fig. 1). The intracellular portion, however, showed less than 20% sequence identity with any other protein of the Swissprot databank. The expression of the novel gene was examined on a Northern blot containing poly(A) RNA from total mouse embryos of four developmental stages (Fig. 2). A radiolabeled cDNA probe for Fgfrl1 hybridized speci¢cally to a mRNA of V2600 nucleotides consistent with the size of the mouse cDNA sequence assuming a poly(A) tail of 300 nucleotides. This mRNA was barely detectable at embryonic day 7, but its signal became clearly visible at day 11 and increased in intensity until day 17. The developmental expression of Fgfrl1 di¡ered signi¢cantly from that of Fgfr1 which was highly expressed at embryonic day 7, but barely expressed at later stages. Hybridization of the same Northern blot with probes speci¢c for mouse
gtgagt
5P UTR, signal
gtcagt
Ig C2
gtaagt
Acidic Box
gtgagt
Ig C2
gtgtgt
Ig C2
(A)n
TM, intracellular, 3P UTR
8166 1108 107 79 243
Fgfr2, Fgfr3, and Fgfr4 did not yield any distinct signals, suggesting very low expression of these genes in total mouse embryos. The mouse cDNA clones were used to screen a genomic DNA library prepared from mouse liver in the vector EMBL3 (Clontech Laboratories ML1030j). After two rounds of screening, a total of nine genomic clones were obtained which together spanned the entire mouse Fgfrl1 gene. Restriction enzyme mapping and direct sequencing of 24 kb of genomic DNA (accession number AJ308490) established the complete exon/intron structure of the mouse gene (Fig. 3). This gene was found to consist of six exons and ¢ve introns. The ¢rst exon which was separated from the rest of the gene by a large intron of 8,2 kb contained the 5P UTR and the sequence for the signal peptide (Table 1). The remaining exons were clustered within 4 kb. Each of these exons coded for a complete, functional domain. Three exons coded for the three IgC2 domains, one coded for the acidic box, and the last exon coded for the transmembrane domain and the short Cterminal tail (Table 1). The ¢rst exon was situated within
Fig. 4. Localization of the Fgfrl1 gene on mouse chromosomes by FISH. The FISH signal is shown at the left, the DAPI banding pattern in the middle. The position of the gene was assigned to region 5E3-F by superimposing the FISH signal and the DAPI banding pattern from ten independent determinations as shown on the right.
BBAEXP 91567 4-9-01
Cyaan Magenta Geel Zwart
250
M. Wiedemann, B. Trueb / Biochimica et Biophysica Acta 1520 (2001) 247^250
a typical CpG island where the dinucleotide sequence CG occurred with high frequency relative to the rest of the genome (Fig. 3). CpG islands are characteristic features of the promoters for housekeeping genes and many proto-oncogenes [11]. When the upstream region of our genomic sequence was carefully inspected, part of the 3P end from the Idua gene was identi¢ed. Hence, the Fgfrl1 gene is separated from the Idua gene by just 10 kb and both genes point into the same direction. The Idua gene is known to encode L-iduronidase and to consist of 14 exons [12,13]. Mutations in this gene are the principal cause of mucopolysaccharidosis type I [14]. Restriction mapping indicated that our genomic clones did not comprise the entire Idua gene, but only exons 11^14. The intervening sequence between the Idua and the Fgfrl1 gene harbored several repetitive DNA elements. There was one copy of the IAPLTR1 MM element (position 2278^2609), one copy of a novel element (position 4899^5045) and two copies of the MMB2 element (positions 6377^6568, 6961^7148). The signi¢cance of these repetitive elements is not known. The location of the novel gene (approved gene symbol Fgfrl1) was mapped on mouse chromosomes by the FISH technique (Fig. 4). A biotinylated restriction fragment of 5.5 kb encompassing exons 2^6 (positions 18 686^24 173) was hybridized to mouse chromosomes prepared from spleen lymphocytes [15,16]. This probe bound speci¢cally to mouse chromosome 5 as demonstrated by staining with £uorescently labeled avidin. Among 100 metaphase spreads analyzed, 87 contained at least two hybridization signals per pair of chromosomes. The exact location as determined by superimposing the FISH signal and the DAPI banding pattern was found to be region E3-F (Fig. 4). This result is consistent with the location of the human FGFRL1 gene that has been mapped to human chromosome 4 band 4p16 [6]. Region 4p16^4q21 is syntenic with region 18^57 cM of mouse chromosome 5, constituting homology group 38 (see Human-Mouse Homology Map at www.ncbi.nlm.nih/Homology). The mouse Idua gene which according to our sequencing results is situated just 10 kb upstream from the Fgfrl1 gene has been placed to chromosome 5 position 57.0 cM by previous genetic analyses. Our mapping data thus allow the correlation of the cytogenetic region 5 E3-F with the genetic region 5 57.0 cM. This region must have undergone substantial rearrangement during evolution of mice and men, because the three human genes FGFRL1, IDUA and FGFR3 are clustered in chromosomal band 4p16, whereas the mouse Fgfrl1 and Idua genes are situated at
BBAEXP 91567 4-9-01
position 5 57 cM, but Fgfr3 is found far apart at position 5 20 cM. Our localization studies might aid in the ¢nal e¡orts to establish a comprehensive map of the mouse genome and to unravel its complete DNA sequence. Our data should also permit the generation of a transgenic animal with a targeted disruption of the Fgfrl1 gene. This study was supported by the Swiss National Science Foundation (Grant 31-61296.00).
References [1] W.L. McKeehan, F. Wang, M. Kan, The heparan sulfate-¢broblast growth factor family : diversity of structure and function, Prog. Nucleic Acid Res. Mol. Biol. 59 (1998) 135^176. [2] G. Szebenyi, J.F. Fallon, Fibroblast growth factors as multifunctional signaling factors, Int. Rev. Cytol. 185 (1999) 45^106. [3] D.M. Ornitz, N. Itoh, Fibroblast growth factors, Genome Biol. 2 (2001) 3005.1^3005.12. [4] L. Pellegrini, D.F. Burke, F. von Delft, B. Mulloy, T.L. Blundell, Crystal structure of ¢broblast growth factor receptor ectodomain bound to ligand and heparin, Nature 407 (2000) 1029^1034. [5] D. Belluoccio, B. Trueb, Matrilin-3 from chicken cartilage, FEBS Lett. 415 (1997) 212^216. [6] M. Wiedemann, B. Trueb, Characterization of a novel protein (FGFRL1) from human cartilage related to FGF receptors, Genomics 69 (2000) 275^279. [7] I. Kim, S.-O. Moon, K.-H. Yu, U.-H. Kim, G.Y. Koh, A novel ¢broblast growth factor receptor-5 preferentially expressed in the pancreas, Biochim. Biophys. Acta 1518 (2001) 152^156. [8] E. Gustafsson, R. Fa«ssler, Insights into extracellular matrix functions from mutant mouse models, Exp. Cell Res. 261 (2000) 52^68. [9] J. Schultz, R.R. Copley, T. Doerks, C.P. Ponting, P. Bork, SMART : a web-based tool for the study of genetically mobile domains, Nucleic Acids Res. 28 (2000) 231^234. [10] G. Von Heijne, A new method for predicting signal sequence cleavage sites, Nucleic Acids Res. 14 (1986) 4683^4690. [11] S.H. Cross, A.P. Bird, CpG islands and genes, Curr. Opin. Genet. Dev. 5 (1995) 309^314. [12] H.S. Scott, X.-H. Guo, J.J. Hopwood, C.P. Morris, Structure and sequence of the human alpha-L-iduronidase gene, Genomics 13 (1992) 1311^1313. [13] L.A. Clarke, J. Nasir, H. Zhang, H. McDonald, D.A. Applegarth, M.R. Hayden, J. Toone, Murine alpha-L-iduronidase: cDNA isolation and expression, Genomics 24 (1994) 311^316. [14] H.S. Scott, S. Bunge, A. Gal, L.A. Clarke, C.P. Morris, J.J. Hopwood, Molecular genetics of mucopolysaccharidosis type I: diagnostic, clinical, and biological implications, Hum. Mutat. 6 (1995) 288^ 302. [15] H.H.Q. Heng, J. Squire, L.-C. Tsui, High-resolution mapping of mammalian genes by in situ hybridization to free chromatin, Proc. Natl. Acad. Sci. USA 89 (1992) 9509^9513. [16] M. Imhof, B. Trueb, Comparative cytogenetic mapping of COL14A1, the gene for human and mouse collagen XIV, Cytogenet. Cell Genet. 84 (1999) 217^219.
Cyaan Magenta Geel Zwart