Portal de Periódicos da CAPES

Secreted Fringe-like Signaling Molecules May Be Glycosyltransferases

1997; Cell Press; Volume: 88; Issue: 1 Linguagem: Inglês

10.1016/s0092-8674(00)81852-8

1097-4172

Yan P. Yuan, Jörg Schultz, Marek Mlodzik, Peer Bork,

Ubiquitin and proteasome pathways

Pattern formation during development requires the regulated expression of numerous signaling molecules. One of these, Drosophila Fringe (FNG) is a novel secreted protein with a key role in dorsal-ventral aspects of wing formation (11Irvine K.D. Wieschaus E. Cell. 1994; 79: 595-606Abstract Full Text PDF PubMed Scopus (305) Google Scholar). Recently, multiple functions have been assumed for a Xenopus homologue (lunatic Fringe, lFNG) including the induction of mesoderm; a complex expression pattern supports this notion (29Wu J.Y. Wen L. Zhang W.-J. Rao Y. Science. 1996; 273: 355-358Crossref PubMed Scopus (57) Google Scholar). Complex and general functions of Fringe-like proteins are also indicated by the presence of two divergent C. elegans and at least six human homologues identifiable by sequence database searches (Figure caption). Thus, conservation patterns within the FNG family emerge that allow the use of sensitive motif and profile searches (for details see3Bork P. Gibson T. Meth. Enzym. 1996; 266: 162-183Crossref PubMed Google Scholar). Indeed, we found significant similarities of FNG-like proteins to Drosophila Brainiac (BRN; 8Goode S. Morgan M. Liang Y.-P. Mahowald A.P. Dev. Biol. 1996; 178: 35-50Crossref PubMed Scopus (64) Google Scholar) and, surprisingly, also to the Lex1 family of biosynthetic galactosyltransferases (Figure 1). BRN has been proposed to be required for proper contact or adhesion between germline and follicle cells (9Goode S. Wright D. Mahowald A.P. Development. 1992; 116: 177-192PubMed Google Scholar, 8Goode S. Morgan M. Liang Y.-P. Mahowald A.P. Dev. Biol. 1996; 178: 35-50Crossref PubMed Scopus (64) Google Scholar). BRN and FNG share several features: i) they are developmentally regulated, secreted signaling molecules without known receptors, ii) they are required during (dorso-ventral) epithelial patterning (11Irvine K.D. Wieschaus E. Cell. 1994; 79: 595-606Abstract Full Text PDF PubMed Scopus (305) Google Scholar, 29Wu J.Y. Wen L. Zhang W.-J. Rao Y. Science. 1996; 273: 355-358Crossref PubMed Scopus (57) Google Scholar, 8Goode S. Morgan M. Liang Y.-P. Mahowald A.P. Dev. Biol. 1996; 178: 35-50Crossref PubMed Scopus (64) Google Scholar), iii) they interact genetically with the Notch and/or EGF receptor pathways (13Kim J. Irvine K.D. Carroll S.B. Cell. 1995; 82: 795-802Abstract Full Text PDF PubMed Scopus (236) Google Scholar; 8Goode S. Morgan M. Liang Y.-P. Mahowald A.P. Dev. Biol. 1996; 178: 35-50Crossref PubMed Scopus (64) Google Scholar), suggesting that they might modify the signaling mediated by these receptors, and iv) FNG and BRN both have at least two C. elegans and several human homologues suggesting the presence of multigene families (Figure 1). Lex1 of Haemophilus influenzae is essential for the biosynthesis of its extracellular lipooligosaccharides (LOS) (5Cope L.D. Yogev R. Mertsola J. Latimer J.L. Hanson M.S. McCracken G.H. Hansen E.J. Mol. Microbiol. 1991; 5: 1113-1124Crossref PubMed Scopus (32) Google Scholar) as is its homologue in another parasitic bacterium, Pasteurella haemolytica (19Potter M.D. Lo R.Y. FEMS Microbiol. Lett. 1995; 129: 75-81PubMed Google Scholar). In two other parasites with a similar LOS architecture, Neisseria meningitidis and Neisseria ghonorrhoeae, two highly related proteins of the Lex1 family have been independently characterized in each organism as galactosyltransferases (10Gotschlich E.C. J. Exp. Med. 1994; 180: 2181-2190Crossref PubMed Scopus (169) Google Scholar, 12Jennings M.P. Hood D.W. Peak I.R.A. Virji M. Moxon E.R. Mol. Microbiol. 1995; 18: 729-740Crossref PubMed Scopus (167) Google Scholar) that add galactose to glucose or N-acetylglucosamine residues of the LOS (12Jennings M.P. Hood D.W. Peak I.R.A. Virji M. Moxon E.R. Mol. Microbiol. 1995; 18: 729-740Crossref PubMed Scopus (167) Google Scholar). The LOS of all these parasitic bacteria contain epitopes that are antigenically and structurally very similar to carbohydrates present in human glycosphingolipids; the parasites are thus able to mimic the latter (16Mandrell R.E. McLaughlin R. Kwaik Y.A. Lesse A. Yamasaki R. Gibson B. Spinola S.M. Apicella M.A. Infect. Immun. 1992; 60: 1322-1328Crossref PubMed Google Scholar). Furthermore, the bacterial galactosyltransferases are significantly similar (blast (1Altschul S.F. Boguski M.S. Gish W. Wootton J.C. Nat. Genet. 1994; 6: 119-129Crossref PubMed Scopus (645) Google Scholar) value for the probability of matching by chance p = 1.9x10−7) to a putative secreted protein from C. elegans (Figure 1) as well as to human and mouse ESTs (blast p-values <10−11). The two most conserved regions of all three subfamilies FNG, BRN, and Lex1 (motif 3 and 4 in Figure 1) are also the major hallmarks of the putative glycosyltransferase superfamily (a more precise prediction of the substrate specificity is not possible given the limited sequence similarity between the subfamilies; 18Paulson J.C. Colley K.J. J. Biol. Chem. 1989; 264: 17615-17618Abstract Full Text PDF PubMed Google Scholar). In the context of an enzymatic function it has to be noted that only negatively charged amino acids are invariant, thus pointing to catalytic residues similar to those in characterized glycosyltransferase families (22Saxena I.M. Brown R.M. Fevre M. Geremia R.A. Henrissat B. J. Bacteriol. 1995; 177: 1419-1424Crossref PubMed Google Scholar, 27Strokopytov B. Penninga D. Rozeboom H.J. Kalk K.H. Dijkhuizen L. Dijkstra B.W. Biochemistry. 1995; 34: 2234-2240Crossref PubMed Scopus (138) Google Scholar, 20Qian M. Haser R. Buisson G. Duee E. Payan F. Biochemistry. 1994; 33: 6284-6294Crossref PubMed Scopus (285) Google Scholar). This is supported by secondary structure predictions (21Rost B. Sander C. Schneider R. Comput. Apll. Biosci. 1994; 10: 53-60PubMed Google Scholar) that are consistent around all conserved regions (Figure 1) and that predict the conserved negatively charged residues to be located in exposed loops indicating a catalytic role. The alternating arrangement of α-helices and β strands suggest an α/β folding type for the central portion of each subfamily, similar to that of other glycosyltransferases (27Strokopytov B. Penninga D. Rozeboom H.J. Kalk K.H. Dijkhuizen L. Dijkstra B.W. Biochemistry. 1995; 34: 2234-2240Crossref PubMed Scopus (138) Google Scholar, 20Qian M. Haser R. Buisson G. Duee E. Payan F. Biochemistry. 1994; 33: 6284-6294Crossref PubMed Scopus (285) Google Scholar, 22Saxena I.M. Brown R.M. Fevre M. Geremia R.A. Henrissat B. J. Bacteriol. 1995; 177: 1419-1424Crossref PubMed Google Scholar). The conclusion that signaling molecules involved in pattern formation such as Fringe and Brainiac, may be secreted glycosyltransferases might come as a surprise, but is not completely unexpected: i) Glycosyltransferases have been implicated in developmental processes for a long time (24Shur B.D. Dev. Biol. 1977; 58: 23-39Crossref PubMed Scopus (55) Google Scholar, 25Shur B.D. Dev. Biol. 1977; 58: 40-55Crossref PubMed Scopus (32) Google Scholar); ii) many extracellular, highly expressed glycosyltransferases have been shown to exist in humans (15Lammers G. Jamieson J.C. Biochem. J. 1988; 256: 623-631Crossref PubMed Scopus (66) Google Scholar, 6Fujita-Yamaguchi Y. Yoshida A. J. Biol. Chem. 1981; 256: 2701-2706Abstract Full Text PDF PubMed Google Scholar); iii) the expression of secreted glycosyltransferases increases during embryonic development (4Cho S.K. Yeh J. Cho M. Cummings R.D. J. Biol. Chem. 1996; 271: 3238-3246Abstract Full Text Full Text PDF PubMed Scopus (48) Google Scholar); iv) the extracellular carbohydrate moieties change during development (17Masteller E.L. Larsen R.D. Carlson L.M. Pickel J.M. Nickoloff B. Lowe J. Thompson C.B. Lee K.P. Development. 1995; 121: 1657-1667PubMed Google Scholar) as a function of the expressed glycosyltransferases (14Kukowska-Latallo J.F. Larsen R.D. Nair R.P. Lowe J.B. Genes Dev. 1990; 4: 1288-1303Crossref PubMed Scopus (471) Google Scholar); and v) transmembrane galactosyltransferases have been shown to transmit intracellular signals after binding substrates via their extracellular part (7Gong X. Dubois D.H. Miller D.J. Shur B.D. Science. 1995; 269: 1718-1721Crossref PubMed Scopus (151) Google Scholar). Secreted glycosyltransferases may use their ability to recognize specific carbohydrate moieties on cell surface molecules to trigger particular receptors (26Shur B.D. Curr. Opin. Cell Biol. 1993; 5: 854-863Crossref PubMed Scopus (93) Google Scholar), but they might also play a crucial role in epithelial pattern formation by modifying these carbohydrate moieties at particular locations recognizable by various carbohydrate-binding domains of extracellular proteins. Numerous distinct ESTs from multicellular organisms including Arabidopsis (no match was found in yeast or other completely sequenced unicellular genomes) suggest a vast superfamily of glycosyltransferases that might belong to a system of posttranslational modification independent from the Golgi apparatus. The carbohydrate status of the cell during development might even be a function of neighboring cells and not only of its own expression set of glycosyltransferases.