O-fucosyltransferase 1/neurotic
The demonstration that O-fut1 is broadly required for Notch signaling raises the question of whether it is simply a permissive factor that is expressed and required ubiquitously, or whether instead its expression might be patterned and contribute to the regulation of Notch activation. The human PO-fut1 gene has been described as being ubiquitously expressed, but its expression has only been analyzed at the level of whole tissue Northern blotting (Wang, 2001). To investigate the expression of Drosophila O-fut1, in situ hybridization was conducted to Drosophila tissues throughout development. O-fut1 mRNA is expressed ubiquitously and strongly in early embryos, from stages one through eight. Since zygotic transcription begins around stage four, the detection of high levels of expression prior to this indicates that O-fut1 mRNA is provided maternally. Indeed, high levels of O-fut1 are detected in germline cells during oogenesis. During germband extension, around stage 9, O-fut1 levels begin to decline, and there is a sharp drop in mRNA levels around stage 11. After stage 12 O-fut1 mRNA expression is greatly diminished in most cells, but it remains highly expressed in a few tissues including the lymph gland and the germ cells of the gonad (Okajima, 2002).
At larval stages, O-fut1 is uniformly expressed in some tissues, such as the wing, leg, and haltere imaginal discs, but in other tissues it is spatially regulated. In the eye disc, O-fut1 mRNA levels are upregulated behind the morphogenetic furrow, and in the larval brain and CNS O-fut1 is expressed in a complex pattern including the proliferation centers of the optic ganglia, and large cells that have been tentatively identified as neuroblasts in other regions (Okajima, 2002).
In order to examine loss-of-function phenotypes for O-fut1, dsRNA-mediated RNAi was employed in flies. dsRNA was generated by using the UAS-Gal4 system to express a portion of O-fut1 in an inverted repeat with an intervening loop (iO-fut1). This expressed dsRNA method was used rather than injecting dsRNA for several reasons, including more consistent phenotypes, greater temporal and spatial control, and the ability to downregulate O-fut1 during larval and adult development. Moreover, because UAS-Gal4 driven expression is temperature-sensitive in Drosophila, and different insertions of the same UAS construct can be expressed at different levels, this method allows for significant control over expression levels. As a test of the effectiveness of the RNAi construct, O-fut1 mRNA levels were monitored by quantitiative, real-time RT-PCR. O-fut1 mRNA levels are reduced to 7% of wild-type levels in whole larvae expressing iO-fut1 ubiquitously under da-Gal4 control (Okajima, 2002).
O-fucose serves as a substrate for the N-acetylglucosaminyltransferase activity of Fng. Hence loss of O-fucose by downregulation of O-fut1 should at least prevent Fng-dependent modulation of Notch signaling. The consequences of downregulation of O-fut1 in the wing were investigated by driving expression of iO-fut1 along the anterior-posterior (A-P) boundary under the control of the patched (ptc) promoter. In ptc-Gal4 UAS-iO-fut1 flies, wing tissue is lost from the distal edge, where the A-P boundary intersects the wing margin. The wing margin is specified by a Fng-dependent activation of the Notch pathway, and loss of tissue from the edge of the wing can also be induced by loss of fng or Notch. ptc-Gal4 UAS-iO-fut1 flies also have phenotypes in other adult tissues that are consistent with a loss of fng-dependent Notch signaling, including leg segment fusions, and at low frequency, small eyes (Okajima, 2002).
To confirm that the phenotypes of iO-fut1 expression result specifically from decreased levels of O-fut1, a UAS-O-fut1 transgene was created. Overexpression of O-fut1 under ptc-Gal4 control in wild-type flies can result in a mild wing vein phenotype. Importantly, however, when UAS-O-fut1 and UAS-iO-fut1 are co-expressed under ptc-Gal4 control, their phenotypes are mutually suppressed, resulting in phenotypically wild-type wings (Okajima, 2002).
To confirm that the loss of wing tissue in ptc-Gal4 UAS-iO-fut1 flies reflects a loss of Notch signaling, the expression of Notch-dependent target genes was examined in late third-instar wing imaginal discs. The expression of cut and wingless (wg), as well as expression driven by the boundary enhancer of vestigial (vg), is detected specifically along the D-V border of the wing imaginal disc, reflecting the peak of Notch activation in the wing. A fourth target, the E(spl)mβ gene, is expressed throughout the wing, presumably because it responds to much lower levels of Notch activation. The vg boundary enhancer and the E(spl)mβ enhancer contain binding sites for Su(H), and appear to be direct targets of Notch signaling. Importantly, the expression of all four Notch targets is repressed along the A-P border in wing discs from ptc-Gal4 UAS-iO-fut1 animals. Thus, reduction of O-fucose levels by downregulation of O-fut1 inhibits Notch signaling during Drosophila wing development (Okajima, 2002).
As a further test of the requirement for O-fut1 in fng-dependent Notch signaling, advantage was taken of the observation that ectopic Fng expression in ventral cells under ptc-Gal4 control induces ectopic Notch activation. The ability of Fng to induce Notch activation in ventral cells is completely suppressed by co-expression with iO-fut1. Since Fng requires an EGF-O-fucose substrate, this suppression also supports the conclusion that O-fut1 activity is significantly impaired in cells expression iO-fut1 (Okajima, 2002).
Because fng serves to position, rather than simply to promote, Notch activation in the wing, loss of fng also results in ectopic Notch activation, which is manifest as the induction of Notch target gene expression and ectopic wing margins. By contrast, ptc-Gal4 UAS-iO-fut1 flies never have ectopic Notch activation. This suggests that O-fucose is not simply a substrate for Fng, but rather that it may have some additional function that prevents ectopic Notch activation despite the absence of Fng-dependent glycans (Okajima, 2002).
In order to investigate other potential requirements for O-fucose during development, UAS-iO-fut1 was crossed to several different Gal4 lines that drive expression in a variety of different tissues and patterns. These resulted in a range of phenotypes including rough eyes, thickened wing veins, loss of notal and abdominal microchaete, induction of additional notal macrochaete, leg segment fusions, and loss of wing margin. When iO-fut1 was expressed broadly (e.g., da-Gal4), high levels of expression caused lethality. All of the phenotypes observed are consistent with a reduction in Notch signaling. Notably, however, many of these phenotypes, such as thickened wing veins or loss of notal microchaete are not associated with mutation of fng. This implies that O-fucose is not simply an acceptor substrate for Fng, but rather is required more generally for Notch signaling (Okajima, 2002).
Notch plays important roles in a wide variety of events, but its effects can be broken down into a few distinct modes. These modes are distinguished both by the way the Notch pathway is utilized and types of modulators that are employed to regulate Notch signaling. The two best studied and most widely utilized Fng-independent modes of Notch signaling are lateral inhibition and cell lineage decisions. During the development of the Drosophila nervous system, Notch-dependent lateral inhibition functions to limit the number of neural precursor cells. In the peripheral nervous system, the sensory organ precursor cells (SOPs) then go through asymmetric cell lineage decisions, each of which is regulated by Notch, to generate the distinct cell types of the sensory organ (Okajima, 2002).
Notal microchaete are mechanosensory organs, and their loss when iO-fut1 is expressed throughout the notum under ap-Gal4 control could be consistent either with increased Notch signaling during lateral inhibition, resulting in loss of SOPs, or decreased Notch signaling during lineage decisions, resulting in loss of the bristle-forming shaft cell. To investigate these possibilities, the influence of iO-fut1 expression on SOP development was examined using neuralized (neur) expression as a molecular marker of SOP fate. An enhancer trap insertion, neur-lacZA101, can be used to detect expression of neur in the SOP, and because of the perdurance of β-galactosidase, in its progeny. Notal microchaete SOPs are specified around 12 hr after puparium formation (APF), but for technical reasons pupal nota were stained at 15-18 hr APF, at which time well-spaced clusters of 2-3 cells were revealed by neur-lacZA101 staining in wild-type animals. By contrast, in ap-Gal4 UAS-iO-fut1 nota, a dramatic increase in neur-lacZA101 staining is observed, and neur-lacZA101 expression is not evenly spaced. Thus, ap-Gal4 UAS-iO-fut1 animals exhibit a failure in lateral inhibition (Okajima, 2002).
To directly examine the influence of O-fucose on cell lineage decisions, markers were used for two of the distinct cell types that are generated from each SOP. ELAV is a nuclear protein that is expressed in neurons, while Prospero is a transcription factor that is expressed in the sheath cell. While the neur-lacZA101 staining described above was conducted on animals raised at 29°C, by 24–30 hr APF, when the sheath and neuron cells have been specified, the nota were too fragile to be stained, presumably because of an extensive loss of epithelial cells. Thus, ap-Gal4 UAS-iO-fut1 animals were raised at 25°C, which generates a weaker phenotype but allowed the tissue to be stained. Strikingly, the number of neurons was significantly increased in these animals, while at the same time the number of sheath cells was significantly decreased. The simultaneous increase in neurons and decrease in sheath cells is diagnostic for a loss of Notch signaling in the presumptive sheath cell during the pIIIb cell division. Although the phenotype is somewhat variable, presumably due to the relatively weak expression of iO-fut1, it was noted that in many cases clusters of cells are observed that contain more than two neurons. This observation is consistent with the hypothesis that O-fut1 is in fact required for other asymmetric cell divisions in the SOP lineage. Similarly, the absence of bristle or socket cells in the adult notum, despite the substantial increase in SOPs, implies that O-fut1 is also required for Notch activation in the presumptive pIIa cell. Together, these results establish that O-fut1 is required for the Notch-dependent processes of lateral inhibition and asymmetric cell lineages during microchaete development, neither of which requires fng (Okajima, 2002).
When iO-fut1 is expressed under ptc-Gal4 control, a sharp boundary exists between iO-fut1 expressing and non-expressing cells. Expressing cells can be simultaneously marked by expression from a UAS-GFP transgene. Importantly, all four of the molecular markers of Notch signaling examined were repressed cell autonomously in ptc-Gal4 UAS-iO-fut1 wing imaginal discs. This cell-autonomous effect indicates that O-fut1 activity is required for the cell's ability to receive ligand signals, and would thus be consistent with the hypothesis that the key substrate of O-fut1 is Notch (Okajima, 2002).
However, the Notch ligands Dl and Ser are also modified by O-fucose. Expression of Dl and Ser in the wing is maintained by Notch activation, hence their transcription decreases when Notch signaling is impaired. Thus, in order to examine the ability of Notch ligands to signal in the presence of iO-fut1, the ligands were expressed under the control of the ptc promoter. This results in strong Notch activation in cells outside the ptc stripe. Inside the ptc stripe, Notch activation is weaker and incomplete because high level expression of Notch ligands inhibits Notch signaling. When Dl or Ser is co-expressed with iO-fut1, Notch activation, as monitored by Wg expression, is suppressed inside the ptc stripe, but not outside of the stripe. Although these experiments are subject to the caveat that O-fucosylation may not be completely inhibited, these results indicate that a reduction in O-fut1 levels that impairs the ability of cells to receive Notch signals does not detectably diminish their ability to send them. They also show that the block in Notch activation associated with reduction of O-fucose is not overcome by elevating expression of Notch ligands (Okajima, 2002).
Activation of Notch involves proteolytic cleavages, and expression of the intracellular domain of Notch alone results in a constitutive, ligand-independent activation of the Notch pathway. If O-fut1 influences Notch signaling by glycosylating its extracellular EGF domains, then the expression of the intracellular domain of Notch should be epistatic to expression of iO-fut1. To test this, both constructs were expressed under ptc-Gal4 control. Indeed, the ability of the intracellular domain of Notch to activate downstream targets is unaffected by co-expression with iO-fut1 (Okajima, 2002).
Whether O-fucose influences the stability or subcellular localization of Notch was also investigated. In wild-type, Notch is expressed ubiquitously throughout third instar wing imaginal disc, with the highest protein levels concentrated apically. In wing cells expressing iO-fut1, Notch is readily detected and the subcellular distribution of Notch is similar to wild-type, with the highest levels of Notch protein observed at the apical and basal edges. This distribution indicates that Notch is not unstable when O-fut1 levels are decreased, nor is it stuck in the ER or Golgi. Rather, it appears to traffic through the secretory pathway to the cell surface. Strikingly however, the levels of Notch protein were substantially increased. The increased level of Notch protein detected is consistent with the hypothesis that O-fucose is required for the activation of Notch -- because Notch is cleaved upon activation, failure to signal to Notch can be expected to result in increased levels of Notch. Indeed, loss of Dl has been reported to cause just such an increase in the Drosophila ovary (Okajima, 2002).
The patterned expression of O-fut1 mRNA suggests that O-fut1 expression levels might be regulated during normal development to influence Notch signaling. In order to begin to examine this possibility, the consequence of ectopic or overexpression of O-fut1 was examined. While lower level expression of O-fut1 generates only the mild wing vein phenotype, higher level expression results in a wide range of phenotypes that are characteristic of a loss of Notch signaling. Thus, depending on the Gal4 line used to drive expression of UAS-O-fut1, and the strength of expression driven by a particular insertion, rough eyes, thickened wing veins, loss of tissue from the wing margin, or extra bristles on the notum or abdomen are observed (Okajima, 2002).
While many of the O-fut1 overexpression phenotypes are similar to those observed when iO-fut1 is expressed, overexpression of O-fut1 results in increased bristle density, instead of bald cuticle. This suggests that increased O-fucosylation impairs lateral inhibition, but not SOP lineage decisions. Consistent with this, staining for expression of ELAV and Prospero at pupal stages reveals an increased density of sensory organs, yet each ELAV-expressing cell remains closely associated with a Prospero-expressing cell. The difference in phenotypes is not simply quantitative, because reduced expression of iO-fut1 never yields a situation in which lateral inhibition is impaired but SOP lineage decisions are not. Instead, as the levels of iO-fut1 expression are lowered, sensory organs with bristle forming shaft cells are recovered in a well-spaced pattern (Okajima, 2002).
Intriguingly, it was also noticed that in flies expressing O-fut1 under ap-Gal4 control, a particularly strong increase in bristle density occurs in reproducible stripes on the notum. This implies that some other factor exists that can modulate the consequences of O-fut1 overexpression. Since Fng is expressed in stripes in the pupal notum, the influence of Fng on the O-fut1 overexpression phenotype was examined by co-expressing Fng and O-fut1. Indeed, the induction of additional microchaete is partially suppressed in the notums of ap-Gal4 UAS-fng UAS-O-fut1 flies (Okajima, 2002).
Therefore, overexpression of O-fut1 suggests that O-fucose, besides being required for activation of Notch, has a second, negative role in Notch signaling. This negative role appears to be qualitatively distinct, as it impairs inductive signaling and lateral inhibition, but not cell lineage decisions. It is speculated that these opposing influences of O-fucosylation could involve positive and negative roles for distinct EGF domains on Notch or its ligands, coupled to different affinities of these EGF domains for O-fut1. As enzyme levels vary, so too would the profile of O-fucosylation on EGF domains. Intriguingly, O-fut1 mRNA is downregulated in most but not all cells during embryonic development, and its levels are also modulated in certain tissues including the eye imaginal disc, larval brain, and larval CNS. The transition between high and low levels of O-fut1 during embryogenesis corresponds to a period of lateral inhibition, when the neural and ectodermal precursors are specified by Notch signaling. Thus, this process may require moderate levels of O-fut1 for effective Notch regulation. Although there are many regulatory inputs into the Notch pathway, the present data suggest that O-fucose may have multiple, distinct roles in Notch signaling, and that regulation of O-fut1 expression may contribute to the pattern of Notch activation during development (Okajima, 2002).
To identify novel genes involved in the Notch pathway, a collection of zygotic lethal mutation with specific maternal effects was screened for mutations associated with a neurogenic phenotype. Such a mutation, 4R6, was identified in a novel gene that has been named neurotic (nti). While nti zygotic mutant embryos have normal nervous systems, embryos derived from nti homozygous germline clones show a strong neurogenic phenotype that is not significantly paternally rescued. Thus, the maternal effect phenotype of nti is similar to the neurogenic phenotypes associated with mutations that are defective in Notch signalling. Next, tests were performed to see whether nti is required for other Notch-dependent developmental events, such as adult wing patterning. In the adult wing, nti mutant clones result in loss of wing margin and vein thickening, phenotypes that are reminiscent of loss of Notch activity. Notch regulates the expression of wingless (wg) at the dorsal and ventral (D/V) compartmental boundary. Like Notch mutant clones, nti mutant clones across the D/V boundary are associated with loss of Wg expression. These phenotypes suggest that nti encodes an essential component of the Notch pathway (Sasamura, 2003).
The nti mutation 4R6 was mapped to 50C-D by complementation tests to deficiency lines, and nucleotide sequence analysis of the mutant chromosome revealed a nonsense mutation in the annotated gene CG12366. The single mutation changes the Lysine at position 133 in the putative ORF of 402 amino acids to a premature stop codon (AAG to TAG), strongly suggesting that 4R6 causes complete loss of CG12366 function. As expected from the nti maternal effect, CG12366 mRNA is uniformly distributed in early embryos but decreases during later stages. To confirm that the nti mutant phenotype is caused by a disruption in CG12366, whether inducible expression of CG12366 could rescue nti mutant cells was tested. Wg expression in the nti clones is rescued in wing imaginal discs by ubiquitous expression of CG12366 (hs-nti). The maternal neurogenic phenotype can also be partially rescued. A hairpin construct (nti-IR) was constructed to express in vivo a dsRNA that would inhibit CG12366 function. Expression of nti-IR using ptc-Gal4 causes wing nicks reminiscent of loss of Notch activity and nti mutant clones, consistent with RNAi-mediated inactivation of nti. This nti-IR expression construct was used to knock down nti function in cultured cells. These results demonstrate that CG12366 encodes nti function (Sasamura, 2003).
Database searches reveal that nti encodes a putative protein with significant homology to human GDP-fucose O-fucosyltransferase 1 (Wang, 2001). O-FucT-1 is an enzyme that adds a carbohydrate, fucose, to serine or threonine residues in the consensus sequence CXXGGS/TC (X is any amino acids) between the second and third conserved cysteines of EGF repeats. Notch and its ligands, Delta and Serrate, have such sequences in their extracellular EGF repeats, and all of them are O-fucosylated (Moloney, 2000; Moloney, 2000b; Panin, 2002). Drosophila O-FucT-1 has been shown to be required for the activation of Notch by its ligands using a RNAi knockdown approach (Okajima, 2002) and O-FucT-1 mutant mice have been shown to have a phenotype similar to that of Presenilins or RBP-Jkappa (Shi, 2003). Thus the current results confirm these previous observations and support the model that O-FucT-1 is an essential component of Notch signalling (Sasamura, 2003).
Fringe (Fng), a modulator of Notch signalling, adds N-acetylglucosamine (GlcNAc) onto the O-fucose moieties of Notch and its ligands. Modification of Notch by Fng, and subsequent glycosylation of Notch, modulates the binding of Delta and Serrate to Notch. This result suggests that nti enables Notch to respond to its ligands by adding O-linked fucose to the Notch EGF repeats. However, this finding appears unexpectedly because no report has ever associated fng mutations with the neurogenic phenotype. Therefore, nti, like Notch, is essential for lateral inhibition during neuroblast segregation, a process for which fng is probably dispensable. Furthermore, the nti mutant phenotype in adult wings is also distinct from that of fng. These results indicate that nti has a function that is independent of fng. Such a fng-independent function was also proposed from the analysis of the neurogenic phenotype induced by RNAi of Nti/O-fut1 in Drosophila bristles (Okajima, 2002; Sasamura, 2003 and references therein).
To define precisely the function of nti in the Notch signalling cascade, the epistatic relationship between nti and other genes involved in Notch signalling was investigated. It was first asked whether the activity of the Notch intracellular domain (NotchICD), which acts as a constitutively active Notch receptor, requires nti. Expression of Wg is absent in nti mutant clones. However, NotchICD can induce Wg expression in nti mutant clones, just as it can in wild-type cells. A similar experiment was carried out with an overexpressed full-length Notch receptor, which in wild type also induces ectopic Wg. Interestingly, in contrast to NotchICD, the full-length Notch cannot induce ectopic Wg expression in nti mutant clones, indicating that nti is required for the activity of full-length Notch but not for its intracellular domain. These interactions are consistent with the idea that Nti modifies the extracellular domain of Notch or its ligands (Sasamura, 2003).
Next, the epistatic relationship between nti and fng was examined. From the structure of the O-linked tetrasaccharide attached to the EGF repeats of Notch and from enzymatic functions of Nti and Fng, it was speculated that Nti activity is a prerequisite for the function of Fng. Ectopic expression of Fng in the ventral compartment of the wing imaginal disc is known to induce Wg ectopically. By contrast, ectopic expression of Fng in nti mutant clones does not induce ectopic expression of Wg, and endogeneous expression of Wg is absent in these clones. These results indicate that Nti is essential for Fng function. These findings are consistent with the RNAi experiments against Nti/O-fut1 (Okajima, 2002; Sasamura, 2003).
Brückner, K., Perez, L., Clausen, H. and Cohen, S. (2000). Glycosyltransferase activity of Fringe modulates Notch-Delta interactions. Nature 406: 411-415. 10935637
Chen, J., Moloney, D. J., and Stanley, P. (2001). Fringe modulation of Jagged1-induced Notch signaling requires the action of ß4galactosyltransferase-1. Proc. Natl. Acad. Sci. 98: 13716-13721. 11707585
Kao, Y. H., Lee, G. F., Wang, Y., Starovasnik, M. A., Kelley, R. F., Spellman, M. W. and Lerner, L. (1999). The effect of O-fucosylation on the first EGF-like domain from human blood coagulation factor VII. Biochemistry 38: 7097-7110. 10353820
Moloney, D. J., Panin, V. M., Johnston, S. H., Chen, J., Shao, L., Wilson, R., Wang, Y., Stanley, P., Irvine, K. D., Haltiwanger, R. S. and Vogt, T. F. (2000a). Fringe is a glycosyltransferase that modifies Notch. Nature 406: 369-375. 10935626
Moloney, D. J., Shair, L. H., Lu, F. M., Xia, J., Locke, R., Matta, K. L. and Haltiwanger, R. S. (2000b). Mammalian Notch1 is modified with two unusual forms of O-linked glycosylation found on epidermal growth factor-like modules. J. Biol. Chem. 275: 9604-9611. 10734111
Oh, S. W., Kingsley, T., Shin, H. H., Zheng, Z., Chen, H. W., Chen, X., Wang, H., Ruan, P., Moody, M., and Hou, S. X. (2003) Genetics 163: 195-201. 12586707
Okajima, T., K. D. (2002). Regulation of Notch signaling by O-linked fucose. Cell 111: 893-904. 12526814
Okajima, T., Xu, A. and Irvine, K. D. (2003). Modulation of notch-ligand binding by protein O-fucosyltransferase 1 and fringe. J. Biol. Chem. 12909620
Panin, V. M., Shao, L., Lei, L., Moloney, D. J., Irvine, K. D. and Haltiwanger, R. S. (2002). Notch ligands are substrates for protein O-fucosyltransferase-1 and Fringe. J. Biol. Chem. 277: 29945-29952. 12036964
Roos, C., Kolmer, M., Mattila, P. and Renkonen, R. (2002). Composition of Drosophila melanogaster proteome involved in fucosylated glycan metabolism. J. Biol. Chem. 277: 3168-3175. 11698403
Sasamura, T., et al. (2003). neurotic, a novel maternal neurogenic gene, encodes an O-fucosyltransferase that is essential for Notch-Delta interactions. Development 130: 4785-4795. 12917292
Sasamura, T., et al. (2007). The O-fucosyltransferase O-fut1 is an extracellular component that is essential for the constitutive endocytic trafficking of Notch in Drosophila. Development 134(7): 1347-56. Medline abstract: 17329366
Shi, S. and Stanley, P. (2003). Protein O-fucosyltransferase 1 is an essential component of Notch signaling pathways. Proc. Natl. Acad. Sci. 100: 5234-5239. 12697902
Wang, Y. and Spellman, M. W. (1998). Purification and characterization of a GDP-fucose:polypeptide fucosyltransferase from Chinese hamster ovary cells. J. Biol. Chem. 273: 8112-8118. 9525914.
Wang, Y., Shao, L., Shi, S., Harris, R. J., Spellman, M. W., Stanley, P. and Haltiwanger, R. S. (2001). Modification of epidermal growth factor-like repeats with O-fucose. Molecular cloning and expression of a novel GDP-fucose protein O-fucosyltransferase. J. Biol. Chem. 276: 40338-40345. 11524432
date revised: 20 September 2003
Home page: The Interactive Fly © 2003 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.