Protein Interactions

Hedgehog movement is regulated through tout velu-dependent synthesis of a heparan sulfate proteoglycan

Tout velu, a homolog of the mammalian EXT tumor supressor gene family, is required for movement of Hh. In vivo evidence is presented that Ttv is involved in heparan sulfate proteoglycan (HSPG) biosynthesis, suggesting that HSPGs control Hh distribution. In contrast to mutants in other HSPG biosynthesis genes, the activity of the HSPG-dependent FGF and Wingless signaling pathways are not affected in ttv mutants. This demonstrates an unexpected level of specificity in the regulation of the distribution of extracellular signals by HSPGs (The, 1999).

Homozygous ttv animals die at the pupal stage, but when maternal and zygotic ttv activities are removed, ttv embryos show absence of naked cuticle and disappearance of both wg and en expression. These phenotypes are reminiscent of hh or wg segment polarity mutants. This segment polarity phenotype represents the ttv null phenotype since the ttvl(2)00681 allele behaves as a genetic null. In addition, no Ttv protein is detected in Western blots prepared from ttvl(2)00681 embryos (The, 1999).

Ttv encodes a type II transmembrane protein, which could be localized at the plasma membrane or in the membranes of the secretory apparatus. To investigate the subcellular localization of Ttv, the Ttv protein was epitope tagged (ttv-myc) because of the inability of the Ttv polyclonal antibody to recognize Ttv in fixed tissues. When expressed under the control of the UAS promoter using the Gal4-UAS system, this ttv-myc gene is able to rescue the segment polarity phenotype of ttv embryos, demonstrating that the myc epitope-tagged protein is functional (The, 1999).

Staining of embryos or imaginal discs expressing ttv-myc construct under control of en-Gal4 revealed a perinuclear and punctate staining. To determine in which subcellular compartments the Ttv protein can be found, whether Ttv-myc colocalizes with either the golgi protein ß-CopII or the endoplasmic reticulum (ER) protein Bip was examined. Interestingly, Ttv-myc mostly colocalizes with Bip and partially with ß-CopII. Further, no plasma membrane staining or a colocalization with the membrane markers E-cadherin or Armadillo was detected. Altogether, these results indicate that Ttv resides mainly in the ER and in the golgi (The, 1999).

In ttv mutant clones induced in the wing imaginal disc, Hh movement and signaling are both reduced. The mammalian Ttv homologs, Ext proteins, have been implicated in HSPG biosynthesis. Therefore, whether HS biosynthesis in Drosophila embryos is affected in the absence of ttv activity was investigated. To detect the presence of HSPGs in vivo, embryos were stained with an antibody (3G10) that has been shown to recognize an epitope on mouse tissues following digestion of HSPGs with heparinase III. Following digestion of the HS sugar chains, one desaturated uronate residue per chain will remain linked to the core protein, enabling the 3G10 antibody to recognize it. Staining of wild-type embryos with 3G10 reveals a uniform staining during early stages and a more pronounced central nervous system staining at later stages. No staining could be detected in embryos that have not been treated with heparinase III (The, 1999).

The 3G10 staining was analyzed in the absence of ttv activity. In contrast to wild type, the staining detected by 3G10 in heparinase-treated ttv embryos was strongly reduced. However, this staining was recovered when wild-type Ttv activity was reintroduced in ttv embryos. These results show that the staining in Drosophila embryos detected by 3G10 is specific and the reduction of staining is due to absence of ttv activity. The reduced staining detected by the 3G10 monoclonal antibody in ttv embryos could not reflect residual Ttv activity since the ttv allele used was a null (The, 1999).

It is possible that other EXT-like genes in Drosophila are responsible for the staining in embryos. Another EXT-like gene, DExt2, has been identified with some homology to ttv. When the DExt2 cDNA clone was sequenced, it was found that it is more homologous to vertebrate EXT2 than to EXT1 (44% and 26% protein identity, respectively). Since DExt2 and ttv are maternally expressed and uniformly distributed in early embryos, it is proposed that the residual 3G10 staining present in ttv embryos is due to the activity of DExt2 (The, 1999).

Since HSPGs are affecting many different factors, such as FGF and Wnts, the specificity of ttv for Hh was investigated. Whether FGF signaling is decreased in the absence of Ttv activity was investigated by analyzing the migration of the mesoderm, a process dependent on the FGF receptor Heartless (Htl) signaling pathway. In wild-type embryos, the ventral mesoderm invaginates at stage 6, and at stage 9 mesoderm cells rearrange and form a monolayer. Mutants with defects in Htl signaling show aberrant mesodermal cell migration, and mesodermal cells do not form a monolayer. Mutations in sgl and sfl exhibit a similar mesoderm migration defect consistent with the role of HSPGs in FGF/Htl signaling. Surprisingly for a mutation involved in HS biosynthesis, no defects in mesoderm migration could be detected in ttv embryos (The, 1999).

ttv embryos display a segment polarity phenotype similar to loss of either Wg or Hh signaling. However, since Wg and Hh signaling pathways in the embryonic epidermis are dependent on each other, the segment polarity phenotype does not allow for distinguishing whether loss of ttv affects Hh as well as Wg signaling. Therefore, two other Wg-dependent processes were examined during embryogenesis; the formation of the stomatogastric nervous system (SNS) neurons and the formation of the RP2 neurons. In wild-type embryos, the invagination of the three SNS neurons can be visualized by staining with antibodies against the Crumbs protein. In mutants that decrease Wg signaling, there are less than three SNS invaginations, while in mutants that increase Wg signaling, more than three invaginations can be found. The SNS phenotype of ttv mutant embryos appears wild type, suggesting that Wg signaling is not affected by loss of Ttv activity. Similarly, no requirement for Ttv activity could be detected in the formation of the RP2 neurons in the embryonic CNS as detected by an Even-skipped (Eve) antibody (The, 1999).

To extend the conclusion that ttv is not required for Wg signaling, another developmental stage was examined: the wing imaginal discs. Wg expressed at the wing margin controls patterning along the D/V axis in a concentration-dependent manner. Expression of the proneural genes at the margin and vestigial (vg) and distalless (dll) at a farther distance are controlled by Wg. In ttv mutant clones at the margin, the expression of the proneural marker A101 is not affected. Further, the expression of dll is not affected. Thus, Wg signaling in wing imaginal discs does not require Ttv activity (The, 1999).

In the absence of sgl or sfl activities, which are also involved in HSPG biosynthesis, both Wg and FGF signaling pathways are reduced. Surprisingly, this analysis of ttv indicates that Hh is specifically affected while both Wg and FGF dependent processes are not altered. This raises the question of whether all HS GAG chains are involved in Hh movement. Therefore, whether Hh signaling is affected in clones of sfl mutant cells in wing imaginal discs was investigated. This analysis suggests that Hh signaling is affected as well in the sfl clones (The, 1999).

The observation that Hh signaling, but not FGF and Wg signaling, is affected in the absence of Ttv activity could indicate that Hh signaling is more sensitive to a reduction of HS GAGs than Wg and FGF signaling. According to this 'quantitative' model, Wg and FGF signaling pathways would not be affected in ttv embryos because HS GAGs synthesized by another Drosophila Ext are sufficient to allow these pathways to function. Alternatively, according to a 'qualitative' model, the specificity of Ttv to Hh signaling suggests the existence of Hh-specific HSPGs. In ttv mutants Wg and FGF, signaling may not be affected because the HSPGs that these factors interact with are present. According to the quantitative model, Wg and FGF signaling pathways would be expected to be at the least partially affected. However, no evidence was found that the activity of these pathways are reduced. Further, in the absence of Ttv activity, the effect on Hh signaling is similar to the loss of Hh activity. Therefore, the 'qualitative' model (i.e., that Ttv activity is required for the synthesis of an Hh-specific HSPG is favored (The, 1999).

Recently, another EXT family member EXTL2, has been identified as an alpha-GlcNAc transferase, which determines that a heparan sulfate instead of a chondroitin sulfate chain will be attached to the linker region of the proteoglycan. The initiation of the GAG chain on the protein core depends on Ser-Gly/Ala dipeptides that have one or more acidic amino acids in close proximity. It has been proposed that the sequences in the core protein surrounding the GAG attachment site are important for the formation of HS chains. Thus, it is possible that different Ext proteins might recognize different sequences on the protein core and thus be specific for certain HSPGs. According to this model, the protein sequence of the HSPG to which Hh binds would be critical for defining its specificity. Another possibility is that Ext proteins generate specific GAG chains, perhaps in a complex with certain HS-modifying enzymes. This model would explain the specificity of ttv on Hh signaling, since ttv would generate a GAG chain specific for Hh (The, 1999).

ttv has been shown to be required for the ability of Hh to reach target cells (Bellaiche, 1998). To extend these observations to embryonic stages, Hh expression was examined in ttv embryos. Staining wild-type embryos with an Hh antibody shows a strong staining in hh-expressing cells and a punctate staining outside of these cells. However, in ttv embryos, Hh is seen only in hh-expressing cells, indicating that Hh does not move beyond its site of production. This phenomenon appears specific to Hh because Wg diffusion is not impaired in ttv embryos (The, 1999).

Hh is produced as a precursor protein, which undergoes autoprocessing. During this process, a cholesterol moeity is attached to the N-terminal portion of Hh (HhNp), which contains the signaling domain. Since HhNp has a cholesterol anchor, it is presumed to remain bound to the membrane. It has been shown that HhN, an N-terminal form of Hh that is not cholesterol modified, can move further in embryos than HhNp and can induce ectopic wg expression. To determine whether the requirement for ttv on Hh diffusion depends on the cholesterol modification of Hh, whether the diffusion of HhN is reduced in ttv embryos was tested by expressing UAS-HhN under the control of en-Gal4. Interestingly, HhN diffuses and induces ectopic en expression. Thus, ttv is required for the proper diffusion of the cholesterol-modified, membrane-associated HhNp but not of unmodified HhN (The, 1999).

These results indicate that HSPGs are involved in the ability of Hh to reach target cells. Hh can act at a distance and is found only at a very low concentration outside Hh-producing cells. Therefore, either the concentration of Hh required to signal is very low and the low amount of diffusible Hh is sufficient for signaling or the membrane tethered Hh can be transported from cell to cell. One model in which HSPGs could influence Hh distribution is by concentrating Hh and perhaps presenting it to its receptor. Such a function has been proposed for HSPGs in FGF signaling. This model assumes that HSPGs are not playing a more active role in the extracellular spreading of Hh (The, 1999).

The observation that membrane-targeted Hh requires HSPGs suggests that there is a transport mechanism for Hh that would allow Hh to move from cell to cell. It is possible that HSPGs are required to target Hh to a specific subcellular compartment. Interestingly, Glypicans with reduced GAG chains are sorted differently from fully glycanated ones and therefore might not able to deliver or aid Hh in the right compartment. The transport of Hh might involve so called 'lipid rafts', which are microdomains in the plasma membrane rich in sphingolipids, cholesterol, and GPI-anchored proteins. Interestingly, Hh has been reported to localize into the detergent-insoluble fraction, characteristic for proteins found in lipid rafts, after separation of cell extracts. Perhaps a GPI-anchored HSPG, such as a Glypican molecule, is required to localize Hh in these rafts. Transfer of GPI-anchored proteins between cells has been observed, and Hh might be transferred from cell to cell in this way. The cholesterol modification on Hh might also facilitate Hh localization into the rafts, after which transport of Hh can occur (The, 1999).

Recently, long cytoplasmic extensions, extending toward the A/P boundary have been identified. Because these filopodia-like structures or cytonemes may transport molecules such as Hh, it has been proposed that the mechanism to generate a morphogen gradient is intracellular and that the time and distance of transport along the cytonemes determines the concentration levels of the signal. If this model proves correct, it is possible that HSPGs are facilitating the transport of Hh down the cytonemes of the receiving cells. It is envisioned that the HSPG could either carry Hh molecules along the membrane or could target Hh to the correct intracellular compartment or vesicles. Interestingly, HhN without cholesterol modification does not require ttv and therefore HSPGs to move from the producing cells. HhN is presumably released into the extracellular matrix, is able to diffuse further than HhNp, and behaves like an ectopic Hh. Therefore, the attachment of HhNp to the membrane and its transport represents another level in the regulation of this potent signaling molecule (The, 1999).

Structural analysis of glycosaminoglycans in Drosophila and Caenorhabditis elegans and demonstration that tout-velu, a Drosophila gene related to EXT tumor suppressors, affects heparan sulfate in vivo

A sensitive method has been devised for the isolation and structural analysis of glycosaminoglycans from two genetically tractable model organisms, the fruit fly, and the nematode. Chondroitin/chondroitin sulfate-derived and heparan sulfate-derived disaccharides were detected in both organisms. Chondroitinase digestion of glycosaminoglycans from adult Drosophila produces both nonsulfated and 4-O-sulfated unsaturated disaccharides, whereas only unsulfated forms are detected in C. elegans. Heparin lyases release disaccharides bearing N-, 2-O-, and 6-O-sulfated species, including mono-, di-, and tri-sulfated forms. Tissue- and stage-specific differences were observed in both chondroitin sulfate and heparan sulfate composition in Drosophila. These methods were applied toward the analysis of tout-velu, an EXT-related gene in Drosophila that controls the tissue distribution of the growth factor Hedgehog. The proteins encoded by the vertebrate tumor suppressor genes EXT1 and 2, show heparan sulfate co-polymerase activity, and it has been proposed that tout-velu affects Hedgehog activity via its role in heparan sulfate biosynthesis. Analysis of total glycosaminoglycans from tout-velu mutant larvae show marked reductions in heparan sulfate but not chondroitin sulfate, consistent with Tout-velu's proposed function as a heparan sulfate co-polymerase (Toyoda, 2000a).

Mutations that disrupt developmental patterning in Drosophila have provided considerable information about growth factor signaling mechanisms. Three genes recently demonstrated to affect signaling by members of the Wnt, transforming growth factor-beta, Hedgehog, and fibroblast growth factor families in Drosophila encode proteins with homology to vertebrate enzymes involved in glycosaminoglycan synthesis. This study reports the biochemical characterization of glycosaminoglycans in Drosophila bearing mutations in sugarless, sulfateless, and tout-velu. Mutations in sugarless, which encodes a protein with homology to UDP-glucose dehydrogenase, compromise the synthesis of both chondroitin and heparan sulfate, as would be predicted from a defect in UDP-glucuronate production. Defects in sulfateless, a gene encoding a protein with similarity to vertebrate N-deacetylase/N-sulfotransferases, do not affect chondroitin sulfate levels or composition but dramatically alter the composition of heparin lyase-released disaccharides. N-, 6-O-, and 2-O-sulfated disaccharides are absent and replaced entirely with an unsulfated disaccharide. A mutation in tout-velu, a gene related to the vertebrate Exostoses 1 heparan sulfate co-polymerase, likewise does not affect chondroitin sulfate synthesis but reduces all forms of heparan sulfate to below the limit of detection. These findings show that sugarless, sulfateless, and tout-velu affect glycosaminoglycan biosynthesis and demonstrate the utility of Drosophila as a model organism for studying the function and biosynthesis of glycosaminoglycans in vivo (Toyoda, 2000b).

Characterization of Drosophila EXT3 (Botv)

Hereditary multiple exostoses gene (EXT) family members encode glycosyltransferases required for heparan sulfate (HS) biosynthesis in humans as well as in Drosophila. A novel Drosophila EXT protein has been identified with a type II transmembrane topology and its glycosyltransferase activities have been demonstrated. The truncated soluble form of this new homolog, designated DEXT3, transfers N-acetylglucosamine (GlcNAc) through an alpha1,4-linkage not only to N-acetylheparosan oligosaccharides that represent growing HS chains (alpha-GlcNAc transferase II activity) but also to GlcUAbeta1-3Galbeta1-O-C(2)H(4)NHCbz, a synthetic substrate for alpha-GlcNAc transferase I that determines and initiates HS biosynthesis. The results suggest that DEXT3 is the ortholog of human EXTL3 and Caenorhabditis elegans rib-2. Semiquantitative reverse transcriptase-PCR analysis has revealed ubiquitous expression of the DEXT3 mRNA. Based on the findings of the present study and those of a study where a fly mutant, deficient in the botv gene, identical to DEXT3, affected HS proteoglycan-mediated developmental signalings, it is suggested that DEXT3 with the revealed glycosyltransferase activities is critically involved in HS formation in Drosophila. These results suggest the essential roles of DEXT3, its human ortholog EXTL3, and the C. elegans ortholog rib-2 in the biosynthesis of heparan sulfate and heparin, if present, in the respective organisms (Kim, 2003).

Cholesterol modification is required for Hh assembly into large punctate structures (LPSs) and its apical targeting in a Disp-dependent manner: LPS apical movement requires Ttv-dependent proteoglycans

Hedgehog family members are secreted proteins involved in numerous patterning mechanisms. Different posttranslational modifications have been shown to modulate Hedgehog biological activity. The role of these modifications in regulating subcellular localization of Hedgehog has been investigated in the Drosophila embryonic epithelium. Cholesterol modification of Hedgehog is responsible for Hedgehog assembly in large punctate structures and apical sorting through the activity of the sterol-sensing domain-containing Dispatched (Disp) protein. Movement of these specialized structures through the cellular field is contingent upon the activity of proteoglycans synthesized by the heparan sulfate polymerase Tout-Velu. Finally, the Hedgehog large punctate structures are necessary only for a subset of Hedgehog target genes across the parasegmental boundary, suggesting that presentation of Hedgehog from different membrane compartments is responsible for Hedgehog's functional diversity in epithelial cells (Gallet, 2003).

Evidence is presented for a functional role of cholesterol modification in the control of Hh subcellular localization in the embryonic epithelium. Cholesterol modification is required for Hh assembly into large punctate structures (LPSs) and its apical targeting in a Disp-dependent manner. Furthermore, LPS apical movement requires Ttv-dependent proteoglycans and this movement is necessary for adjacent anterior cells to receive Hh input and express wg. In contrast, basolateral Hh localization is sufficient for rho activation in adjacent posterior cells, independent of cholesterol modification of Hh (Gallet, 2003).

In certain developmental processes, palmitoylation increases Hh activity to reach the threshold necessary for target gene expression. Expressing C85S-Hh-Np lacking palmitic acid modification can rescue hh loss-of-function during embryogenesis, though less efficiently than a wild-type Hh molecule. Hence, as in vertebrates, palmitic acid modification potentiates Hh activity in Drosophila embryos. Interestingly, Hh-N is rendered inactive if not palmitoylated, suggesting that palmitoylation occurs independently of cholesterol modification (Gallet, 2003).

Hh is localized at the basolateral membrane of producing cells. Hh is also present in LPS, the formation of which is cholesterol dependent. Since no difference is observed in LPS formation with a nonpalmitoylated C85S-Hh-Np construct, the cholesterol modification on Hh appears to be the main requirement for Hh targeting to LPSs. Two different fractions of membrane-bound Hh have been identified in Drosophila: a detergent-insoluble fraction corresponding to lipid raft microdomains and a detergent-soluble fraction. Therefore, a potential hypothesis would be that the Hh LPSs correspond to cholesterol-enriched raft microdomains. Nevertheless, it was not possible to show any colocalization of Hh with FloDm, the fly homolog of raft-associated caveolin. It was also not possible to see any Hh-related cuticle defects in embryos injected with drugs (methyl-ß-cyclodextrin and filipin) known to deplete cholesterol and, thus, lipid raft microdomains (Gallet, 2003).

The assembly and the apical sorting of Hh-Np LPSs are dependent upon both cholesterol and Disp activity. However, Disp is not necessary for cholesterol binding to Hh. In C. elegans, the disp homolog (CHE-14) is required for apical cuticle secretion. Therefore, one tempting possibility might be that the sterol-sensing domain (SSD) on Disp specifically recognizes cholesterol-modified Hh for its assembly into LPSs and apical sorting. Since Hh-Np is still present on basolateral membranes in disp mutants, formation of LPSs could start from these locations before apical sorting. Alternatively, two independent routes could be responsible for basolateral and apical targeting (Gallet, 2003).

Cholesterol-dependent Hh-Np LPSs require Ttv to diffuse in the cellular field. How can Hh-Np be released from cells if it is inserted in the lipid bilayers? The mechanism of release might involve either a displacement of the cholesterol tether on Hh-Np or the formation of membrane vesicles. So far, no evidence for vesiculation has been reported, but large soluble multimers of Shh-Np have been identified in conditioned media of vertebrate cells. Multimers of Hh-Np are also present in conditioned culture media from Hh-producing Drosophila Schneider cells. Furthermore, the fact that all Shh or Hh soluble molecules identified so far are cholesterol-modified, strongly suggests that Hh-Np cannot be released from its anchor by cleavage (Gallet, 2003).

How do Hh-Np LPSs move within the cellular field? At least two alternative mechanisms could explain this movement. Planar transcytosis and, thus, transit from cell to cell in an endocytic and recycling-dependent manner might be involved. Alternatively, Hh-Np LPSs could pinch off from membrane raft domains, spread in the extracellular space, and become internalized away from the source at different cell positions. The role of the Ttv-dependent heparan sulfates could either be to stabilize such structures or to transport them from cell to cell (Gallet, 2003).

Differential activation of wg and ptc in anterior cells and of rho and ptc in posterior cells is related to the membrane localization of Hh. Cholesterol-dependent LPS formation and apical targeting are shown to be necessary for proper anterior wg activation but dispensable for rho expression in posterior cells. Conversely, basolateral targeting of Hh in cells producing Hh-N-CD2 and Hh-N-GPI is sufficient to activate the posterior rho expression, independent of the presence of cholesterol (Gallet, 2003).

Interestingly, wg is expressed in adjacent cells located just anterior to the Hh-sending cells. Hence, long-range diffusion of Hh should not be required for wg activation. However, in the absence of Ttv function, Hh-Np LPSs are blocked apically in producing cells, and wg is not activated. Ttv-dependent heparan sulfate proteoglycans are required for long-range Hh-Np movement in the wing disc. Thus, these results suggest that, in the embryonic ectoderm, two different mechanisms of Hh pathway activation are present. wg activation requires all the events previously associated with long-range Hh target gene activation and thus depends on Hh secretion and transport mechanisms. However, rho does not require secretion of Hh and can be activated in a cell-cell contact-dependent manner, like a short-range target. This difference could be due to differential accessibility of Hh to anterior versus posterior cells caused by the presence of the parasegmental boundary between en and wg cells. Indeed, when Ttv is expressed exclusively in cells anterior to En cells, both wg- and rho-dependent cell differentiation are rescued. This indicates that a differential transport and/or presentation of Hh-Np could be responsible for the asymmetric cellular response to Hh (Gallet, 2003).

How then is rho activated in cells posterior to Hh-producing cells? rho expression could depend on cell-cell contact activation with or without internalization of Hh. Although no detectable Hh in Hh-N-CD2 neighboring cells was observed, the possibility cannot be excluded that rho activation might depend on Hh internalization. It is worth mentioning that an Shh-CD4 transmembrane fusion protein has been shown to be internalized in adjacent cells through Ptc-1 activity in mammalian tissue culture cells and can induce formation of the most posterior digit of the chick limb. Moreover, expression of Hh-Np in disp mutant embryos that are defective in apical sorting induces rho expression in several rows of cells. In these embryos small dots of Hh-Np are seen outside the producing cells, confirming a possible internalization of Hh in posterior receiving cells through basolateral membrane interactions. This internalization could propagate at long range, since rho and ptc are activated in six to seven rows of cells when non-cholesterol-modified Hh-N is expressed in disp mutant embryos (Gallet, 2003).

In summary, these data suggests that some Hh/Shh targets can be activated through Hh trafficking in LPSs followed by apical secretion, whereas other targets might be activated by basolaterally targeted Hh. Hence, it is hypothesized that presentation of Hh from different cellular membrane compartments allows the receiving cells to differentially respond to the Hh input. This provides an interesting new paradigm regarding the mode of action of morphogens in all metazoans (Gallet, 2003).

Biochemical studies of Drosophila EXT proteins

To understand the molecular mechanisms by which the Drosophila EXT proteins play distinct and collaborative roles in cell signalling, biochemical experiments were performed to analyze their interactions and subcellular localization. Ttv and Sotv physically associate with each other to form a complex, and they have virtually identical subcellular localizations. However, neither Ttv nor Sotv physically interact with Botv. Botv also has a more diffusive staining in cells than Ttv and Sotv. Consistent with these results, biochemical studies in vertebrates show that vertebrate EXT1 and EXT2 also physically associate with each other to form a complex. Biochemical studies have further demonstrated that both EXT1 and EXT2 have GlcNAc and GlcA transferase activities when expressed independently, although EXT1 has a more robust activity than does EXT2 (Lind, 1998; McCormick, 2000; McCormick, 1998; Senay, 2000; Wei, 2000). However, co-expression of EXT1 and EXT2 has a synergistic effect on enzyme activities (McCormick, 2000; Senay, 2000). These results led to a proposal that the fully active HS co-polymerase may be a complex containing EXT1 and EXT2, in which both subunits are essential for the activity (Zak, 2002). The functions of EXT-like proteins have also been investigated. EXTL3 appears to be a bifunctional alpha GlcNAc transferase that can transfer GlcNAc to both the linkage region and to intermediates during chain polymerization, suggesting that EXTL3 is involved in both the initiation and polymerization of HS GAG chains (Kim, 2001). Importantly, a recent biochemical study demonstrated that Botv has alphaGlcNAc transferase activities (Kim, 2002) that can transfer GlcNAc to both the linkage region and to intermediates in chain polymerization (Han, 2004b).

Heparan sulfate polymerization in Drosophila: Ttv has GlcNAcT-II and GlcAT-II activities required for the biosynthesis of repeating disaccharide units of the HS backbone

The formation of heparan sulfate (HS) chains is catalyzed by glycosyltransferases encoded by EXT (hereditary multiple exostosin gene) family members. Genetic screening for mutations affecting morphogen signaling pathways in Drosophila has identified three genes, tout-velu (ttv), sister of tout-velu (sotv), and brother of toutvelu (botv), which encode homologues of human EXT1, EXT2, and EXTL3, respectively. So far, in vitro glycosyltransferase activities have been demonstrated only for Botv/DEXTL3, which harbors both N-acetylglucosaminyltransferase-I (GlcNAcT-I) and N-acetylglucosaminyltransferase-II (GlcNAcT-II) activities responsible for the chain initiation and elongation of HS, and no glucuronyltransferase-II (GlcAT-II) activity. This study demonstrates that Ttv/DEXT1 and Sotv/DEXT2 have GlcNAcT-II and GlcAT-II activities required for the biosynthesis of repeating disaccharide units of the HS backbone, and the coexpression of Ttv with Sotv markedly augment both glycosyltransferase activities when compared with the expression of Ttv or Sotv alone. Moreover, the polymerization of HS was demonstrated on a linkage region analogue as an acceptor substrate by Botv and an enzyme complex composed of Ttv and Sotv (Ttv-Sotv). In contrast to human, Drosophila Ttv-Sotv exhibit no GlcNAcT-I activity, indicating that Botv/DEXT3, which is an EXT-Like gene and possesses GlcNAcT-I activity required for the initiation of HS, is indispensable for the biosynthesis of HS chains in Drosophila. Thus, all three EXT members in Drosophila, Ttv, Sotv, and Botv, are required for the biosynthesis of full-length HS in Drosophila (Izumikawa, 2006).

EXT genes have been well conserved between Drosophila and humans. So far the Drosophila EXT family genes,ttv, sotv, and botv, which encode Drosophila homologues of human EXT1, EXT2, and EXTL3, respectively, have been identified. This study demonstrates that the polymerization of HS in Drosophila is achieved by an enzyme complex composed of Ttv and Sotv (Ttv-Sotv), analogous to findings made for the human EXT1 and EXT2 protein complex. In contrast to human EXT1-EXT2, which possesses GlcNAcT-I activity for chain initiation, Ttv-Sotv exhibits no GlcNAcT-I activity, indicating that Botv/DEXT3, which is an EXT-Like gene and possesses GlcNAcT-I activity, is indispensable for the biosynthesis of HS chains in Drosophila. Thus, all three EXT members in Drosophila, Ttv, Sotv, and Botv, appear to be essential for the biosynthesis of HS (Izumikawa, 2006).

An indispensable role for Ttv, Sotv, or Botv in HS biosynthesis has also been suggested by the observations that the synthesis is diminished or abolished in Drosophila bearing a mutation in ttv, sotv, or botv. In addition, mutations in ttv, sotv, and botv seriously impair Hh, Wg, and Dpp signaling activities as well as the distribution of these morphogens. Thus, these results also indicate that HS is an important regulator of morphogen signaling and the tissue distribution of these morphogens. Notably, botv null embryos exhibited stronger segment polarity phenotypes than sotv null embryos. These results suggest that a mutation in botv led to impaired HS biosynthesis because the mutants do not have GlcNAcT-I activity for chain initiation on the linkage region. Also, residual staining with a monoclonal antibody 3G10 has been repored in ttv embryos. These reports are consistent with the finding that HS polymerization occurs when Botv and Ttv or Sotv were used as an enzyme source, although the efficiency of polymerization is low. Moreover, Wg signaling is defective only in the botv mutant or ttv-sotv double mutant but not in the ttv or sotv mutant, whereas the distribution of the Wg morphogen is abnormal in the ttv, sotv, and botv mutants. Despite the existence of a small amount of HS in the ttv or sotv mutant, signaling events performed by Wg are abrogated. These results suggest that there might be a threshold amount of HS for the reception of signaling molecules (Izumikawa, 2006).

In mammals, the polymerization of HS occurs with GlcNAc and GlcUA transferred alternately by the action of an enzyme complex consisting of EXT1 and EXT2 from the EXT gene family. These genes are involved in hereditary multiple exostosis, which is an autosomal dominant disorder characterized by the formation of a cartilage-capped tumor, caused by mutations in either EXT1 or EXT2. EXT1- and EXT2-deficient mice generated by gene targeting fail to undergo gastrulation and die by embryonic day 8.5. Moreover, embryonic stem cells from EXT1- and EXT2-deficient mice showed a complete loss of HS. These results indicate that EXT1 and EXT2 are essential for both gastrulation and the biosynthesis of HS early in embryonic development. Recently, it has been reported that EXT1 mutant mice generated by the gene trap method (EXT1Gt/Gt) survive to embryonic day 14.5. The embryonic fibroblasts from EXT-deficient mice generated by the gene targeting method still produce short HS chains as compared with those from wild-type mice. It is suggested that because EXT1 gene trap mouse fibroblasts produce small amounts of normal EXT1 transcript, gene trap mice still produce some HS, albeit much less than normal. Thus, the EXT1 gene trap mice survive longer than the EXT1-deficient mice. These results suggest, as in the case of Drosophila, that the amount or length of HS chains is the critical determinant of embryonic mouse development, and the function of HS is altered by the amount or chain length (Izumikawa, 2006).

The family of mammalian EXT genes has been extended by the identification of three EXT-Like genes, EXTL1, EXTL2, and EXTL3. Although the EXT-Like genes are predicted to be involved in the synthesis of HS on the basis of their enzymatic activities in vitro, the functions of these genes in vivo have not been demonstrated. In fact, the polymerization of HS on the linkage analogue is achieved by an enzyme complex of EXT1 and EXT2, without the aid of EXTL proteins. Most interestingly, the HS chains synthesized on the linkage analogues by the EXT1-EXT2 complex are much longer than those from naturally occurring HS. Thus, it is predicted that these EXTL proteins might regulate the size of HS chains in vivo (Izumikawa, 2006).

tout-velu: Biological Overview | Evolutionary Homologs | Developmental Biology | Effects of Mutation | References

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.