nemo: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References
Gene name - nemo

Synonyms -

Cytological map position - 66A22-B5

Function - signaling

Keywords - epithelial planar polarity,
wing shape, vein patterning

Symbol - nmo

FlyBase ID: FBgn0011817

Genetic map position - 3-

Classification - protein serine/threonine kinase

Cellular location - cytoplasmic and possibly nuclear

NCBI links: Precomputed BLAST | Entrez Gene

Drosophila nemo was originally identified as a tissue-specific factor required for the establishment of epithelial planar polarity (EPP) in developing ommatidia (Choi, 1994). It was subsequently shown to act downstream of Frizzled signaling in the polarity pathway (Strutt, 1997). Nemo is the founding member of a novel family of serine/threonine protein kinases most similar to MAPKs (Choi, 1994). Homologs have been identified in humans (Dres16), mice (Nemo-like kinase, Nlk) and C. elegans (Lit-1) (Banfi, 1996; Brott, 1998; Meneghini, 1999; Rocheleau, 1999; Verheyen, 2001 and references therein).

Recent studies in C. elegans and vertebrates have uncovered functional roles for Nemo homologs in several Wnt-dependent processes (Rocheleau, 1999; Meneghini, 1999; Ishitani, 1999). Biochemical evidence points to an involvement in canonical Wnt signaling. This Wnt pathway (see Drosophila Wingless) is activated when secreted Wnt ligand binds to a Frizzled-type cell surface receptor, which transduces the signal via the cytoplasmic Dishevelled protein. Wnt signaling in both vertebrates and Drosophila acts to stabilize cytoplasmic ß-catenin/Armadillo (Arm), allowing Arm to translocate to the nucleus where it interacts with a member of the TCF/LEF family of transcription factors (Drosophila homolog: Pangolin), and subsequently controls gene expression (Verheyen, 2001 and references therein).

Homologs of Nemo have roles in regulating Wnt-dependent gene expression through their affects on ß-catenin and TCF homologs. Nlk can phosphorylate TCF and inhibit DNA-binding of the ß-catenin/TCF complex, thereby inhibiting Wnt signaling (Ishitani, 1999). In addition, worm LIT-1 can bind to and phosphorylate WRM-1, a ß-catenin/Armadillo-like protein (Rocheleau, 1999). This interaction leads to phosphorylation of POP-1, and its subsequent export from the nucleus. In C. elegans, interactions between lit-1 and two distinct Wnt signaling pathways, one regulating polarity and one regulating gene expression, have been described (Rocheleau, 1999). In addition, vertebrate Nlk has also been shown to inhibit ß-catenin/Arm function in vivo. Injection of Nlk into Xenopus embryos can suppress the axis duplication phenotype induced by microinjected ß-catenin/Arm (Ishitani, 1999; Verheyen, 2001 and references therein).

Drosophila Nemo plays a role in epithelial plane polarity in the wing and abdomen suggesting it may have a more global role in polarity specification. In addition, nmo is required during wing development for vein patterning, wing shape and morphology. Genetic interactions between nmo and mutations in Wg, Notch and Dpp signaling implicate nmo as an antagonist of Wg. nmo mutations also cause numerous pleiotropic effects during development and the gene appears to be essential during embryogenesis (Verheyen, 2001).

Wings of nmo mutants are rounder and shorter than wildtype, contain extra vein material and are held away from the body at a 45° angle (Choi, 1994). A genetic second-site modifier screen was carried out to identify genes that either participate in Notch signaling or modulate cross-talk between signal transduction pathways. adirondack (adk) mutations, now known to be alleles of nmo, dominantly modify the rough eye phenotype caused by expression of activated Notch under expression of the sevenless promoter. The extra vein phenotype seen in hypomorphic nmo mutant alleles (nmoadk) suggests that the gene product normally functions in the suppression of the vein fate. The patterning of wing veins has been well characterized and involves many genes at various stages of specification, refinement and maintenance of vein fates. The roles of the EGF receptor and Notch pathways in these processes are well-characterized. One of the earliest markers for presumptive vein cells is the rhomboid gene. rhomboid facilitates signaling via the EGF receptor and its expression is negatively regulated by Notch. To examine the development of the ectopic veins in nmoadk, pupal wings were dissected and the pattern of Rhomboid mRNA expression was determined. In nmoadk pupal wings ectopic rhomboid expression is seen in the regions corresponding to where ectopic veins are found in adult wings. These results suggest that nmo normally acts to inhibit rho expression (Verheyen, 2001).

The expression and localization of nmo mRNA was also examined in pupal wings, in order to characterize Nemo's role in vein development. nmo transcript is detected in the intervein regions of the wing blade and enriched in the cells juxtaposed to vein-forming cells. Consistent with the results of the Northern blot, in nmoadk mutant wings, nmo transcript is still detected although there are regions where the transcript is greatly reduced. These regions correspond exactly to the sites of ectopic vein formation and ectopic rhomboid expression. These findings support the model of a role for nmo in suppression of vein fate, with localized loss of nmo expression leading to formation of ectopic veins (Verheyen, 2001).

Expression of UAS-Nemo in a nmoadk background suppresses the extra veins seen emanating from the posterior cross-vein (PCV), as well as those near longitudinal veins II and V (LII and LV). Furthermore, in a wildtype background, ectopic expression in intervein cells of UAS-Nemo using 32B-Gal4 and 69B-Gal4 causes loss of posterior cross-vein tissue. These effects depended greatly on the Gal4 driver, suggesting that the precise timing of overexpression is critical to suppression of veins. Interestingly, in the anterior regions of the wing, overexpression of Nemo leads to ectopic vein formation. Ectopic expression of nmo also results in a change in the overall shape of the wing blade. nmoadk mutant wings are much rounder and shorter than normal. In contrast, ectopic nmo causes the wing blade to elongate and become more narrow and pointed (Verheyen, 2001).

The murine Nemo homolog Nlk has been implicated in regulating Wnt signaling by repressing the Arm/ß-Catenin-TCF complex (Ishitani, 1999), a component of Wingless signaling. Whether such an interaction occurs in flies during wing vein formation was examined. The extra vein phenotypes observed in nmo mutant wings are very similar to those that have been previously described as resulting from overexpression of Armadillo and both vertebrate ß-catenin and plakoglobin in the wing. In addition, ectopic veins are produced as a result of ectopic wg and dsh expression. Constitutively active Armadillo (UAS-Arms10) expressed using the 1348-Gal4 driver leads to the formation of moderate ectopic veins emanating from the PCV, in addition to more severe ectopic veins along LII. These are both regions of the wing that are sensitive to nmo mutations and where similar ectopic veins are observed in nmoadk. The phenotypes seen with overexpression of wg and arm are consistent with the theory that Nemo is a negative regulator of Wingless signaling since loss of nemo mimics extra veins seen with overexpression of arm and wg (Verheyen, 2001).

Overexpression of both wildtype mouse Lef-1 (a TCF homolog) and a constitutive repressor form of dTCF (Pangolin) results in dominant negative phenotypes. The constitutive repressor form of dTCF (UAS-dTCFDeltaN) is unable to bind Arm and represses wg-dependent gene expression. These findings suggest that expression of wildtype dTCF (UAS-dTCFwt) somehow interferes with a wingless-targeted transcription factor in a dominant negative way. Consistent with this, it is found that ectopic expression of UAS-dTCFwt using vestigial-Gal4 results in defects in the posterior wing margin, a phenotype seen with loss of wg signaling. Ectopic expression of the UAS-dTCFDeltaN using the 1348-Gal4 driver is lethal. However, homozygosity for nmoadk is able to rescue the lethality and the flies that emerged had reduced ectopic wing veins. This finding can be interpreted by taking into account the dual roles TCF plays in the nucleus. In nmoadk mutants, the negative regulation of endogenous dTCF may be reduced, leading to more wg-dependent signaling and the induction of extra veins similar to those seen with constitutive Arm expression. Expression of UAS-dTCFDeltaN in the nmoadk background most likely interferes with the de-repressed endogenous dTCF and block the induction of extra veins seen in nmoadk (Verheyen, 2001).

Since the eye defect in nmo is caused by disrupted polarity of the ommatidial clusters, it was of interest to examine whether other planar polarity processes are disrupted in nmoadk. The polarity of the wing hairs was examined in nmoP, nmoadk1 and nmoadk2 flies. Normally, each epithelial cell in the wing blade reorganizes its cytoskeleton and produces an actin spike at the distal vertex, which gives rise to the distally oriented wing hairs seen in adults. nmoP flies do not display any noticeable polarity defects in the wing. A polarity defect is observed in nmoadk2 wings in those regions where ectopic veins are seen. A milder, variable polarity defect is observed in nmoadk1 homozygotes. Interestingly, polarity defects are also observed in flies overexpressing UAS-Nemo using the 1348-Gal4 driver. This finding is consistent with studies in which both overexpression and loss-of-function of fz led to polarity defects during eye and wing development. These results have been interpreted to reflect a finely balanced signaling pathway that is easily disturbed by reduced or excessive expression of its components (Verheyen, 2001 and references therein).

Expression of UAS-Nemo using the 69B-gal4 driver results in a number of abnormalities, among them severe patterning defects in the abdomen, including disrupted polarity of microchaete bristles in the tergites. Since other epithelial planar polarity (EPP) mutants also show abdominal phenotypes, it is suggested that Nemo is a part of a conserved signaling mechanism that acts in the establishment of EPP. In addition to polarity disruptions, reduced numbers of bristles and defects in the patterning of the tergites are also seen, most noticeable in the reduction of the pigment band width, with occasional loss of cuticular tissue (Verheyen, 2001).

These results and those of Choi and Benzer (1994) suggest that nmo plays a general role in establishing EPP. Through the analysis of additional alleles and ectopic expression studies, it has been determined that nmo also plays a role in polarity of wing hairs and abdominal bristles. Alterations of signaling by Fz pathway components often leads to unexpected phenotypes, such as the finding that both loss-of-function mutations and overexpression of fz or dsh result in similar phenotypes during eye development. Similarly, in the wing, over-expression of fz leads to a tissue polarity defect that is similar to the loss of function fz phenotype. The finding that both loss of nmo and overexpression of UAS-Nemo cause polarity defects in the wing is consistent with other members of the EPP pathway. The regulation of nmo expression has been shown to be dependent of fz signaling, and it is possible that nmo mutants display aberrant polarity phenotypes as a result of losing some aspect of the downstream events triggered by fz signaling. However, the possibility that Nemo is causing polarity defects through an indirect effect on the regulation of cytoskeletal reorganization cannot be excluded. It is also possible that Nemo's effect on polarity is more indirect and reflects the earlier role for Wg signaling upstream of EPP establishment (Verheyen, 2001).

The nature of the nmoadk phenotype is somewhat complex, given that it appears stronger than the published null allele, nmoP (Choi, 1994). It is speculated that nmoP alleles display the zygotic null phenotype, but that the abundant maternally supplied Nemo permits development through to adult stages. nmoadk alleles appear more severe by several criteria such as their enhanced affect on wing veins, wing blade angle, fertility and viability. The presence of truncated mRNAs provides a possible mechanism. The mRNAs may encode truncated proteins that have a dominant negative effect on the maternal Nemo and may therefore interfere with development at an earlier stage. The elucidation of the phenotype of a maternal and zygotic null should clarify this issue. It is predicted that embryos both maternally and zygotically null for nmo would be embryonic lethal (Verheyen, 2001).

The role of Nemo in the wing seems quite complex and may reflect interactions between nmo and numerous signaling pathways. The role of nmo in wing vein development is reminiscent of the roles played by the other vein-inhibiting genes such as net and plexus. Both the expression pattern of nmo and the ectopic veins seen in the mutant support a role for nmo in promoting intervein cell fates (Verheyen, 2001).

A relationship is observed between overall wing shape and wing vein patterning. nmoadk mutants have excess veins and a much rounder, shorter wing. Such a wing shape has been seen in other excess vein genotypes, such as UAS-Argos and extreme px alleles. Overexpressed Nemo results in a longer and more pointed wing shape, in addition to vein loss in the posterior compartment. Such a wing shape is also seen in mutants that reduce vein formation such as vein; in mutants in TGFß signaling such as dpp and gbb-60A, and in flies ectopically expressing dominant negative Raf. This correlation between excess vein and rounder wing shape and loss of veins accompanied by a narrower wing suggests a link between the genes controlling vein/intervein fates and those regulating overall wing morphology. In addition, it has been shown that mutations that have severe effects on vein formation such as double mutant combinations of extra vein mutants result in a larger wing blade. It is therefore interesting that the net1;nmoadk2 flies have much smaller wing blades. This phenotype is most likely due to the observed vast expanse of vein tissue, which is known to have a smaller apical surface than intervein tissue (Verheyen, 2001 and references therein).

What is the relation of nemo to the Notch pathway? nemo was originally shown to play a role in planar polarity in the developing eye (Choi, 1994). nmo is required for the rotation of photoreceptor clusters in the final stages of polarity establishment. It has been proposed that, in the eye, nmo functions as a tissue-specific factor downstream of the Fz pathway. The initial step in the establishment of polarity in each ommatidium is the differentiation of the R3/R4 photoreceptors. Recent studies have shown that Notch signaling acts downstream of Fz in this process. Fz signaling leads to expression of the Delta ligand in the R3 cell and through a negative feedback loop Notch signaling is restricted to the R4 cell. In this way, Fz signaling can exert control over Notch activity, by initially biasing the ligand production to one cell type (Verheyen, 2001 and references therein).

In the wing, ectopic activation of Notch signaling can suppress the extra veins in nmo and mutations in Notch pathway components enhance the extra veins. Notch signaling is also required for vein differentiation, and genetic interactions suggest that nmo acts upstream of Notch and may participate in laying out the initial vein/intervein boundaries. The ectopic veins seen in nmo resemble those seen in other extra vein mutants, and seem to mark the boundaries of para-vein territories. Ancestors of Drosophila possess veins in these para-vein regions, but their formation has been suppressed in Drosophila (Verheyen, 2001).

The specification of wing veins has been shown to involve a number of signal transduction cascades. Roles for Egfr, Notch and Dpp signaling are well established. More recently, a role for Wg signaling in this process has been suggested. The exact molecular nature of Wg signaling in wing vein specification is unknown, but it has been shown that ectopic signaling through Wingless and Armadillo results in the formation of ectopic veins. Dpp signaling also plays a role in specifying vein fates during pupal stages. This is in contrast to the known role for Notch signaling in inhibiting vein fates. A recent study finds that Egfr signaling is regulated developmentally, such that it stimulates veins early in wing development and later inhibits vein formation. Larval expression of dominant negative Ras signaling results in loss of veins but pupal expression of the same construct resulted in ectopic veins. This finding is intriguing in light of the dual effects on vein specification seen with ectopic expression of UAS-Nemo in the wing blade. Whether this effect occurs as a direct consequence of Nemo on the Egfr pathway, through its control of rhomboid expression, or indirectly through the effect of Nemo on Notch or Wg signaling has yet to be determined. The behavior of nmo in the wing is consistent with its playing a positive role upstream of Notch signaling and acting as an antagonist of Wingless signaling, although the possibility cannot be excluded that the effect of nmo on wg may be either positive or negative depending on the developmental stage. It is possible that through repression of Wg signaling that Nemo facilitates Notch signaling. Both genetic and biochemical cross-regulation between Notch and Wg through the Dsh protein have been described (Verheyen, 2001 and references therein).

Wingless signaling has been linked both molecularly and genetically with the Notch, MAPK, Hedgehog, Dpp and Fz planar polarity signal transduction pathways. Such findings have led to the model that Wnt signaling functions as a network, rather than a linear pathway. Such a network could integrate the cross-regulation occurring between numerous developmentally regulated pathways. The network of interactions must be carefully orchestrated, since shifting the levels of any one protein can tip the balance between developmental pathways. Maintaining the equilibrium between differentially activated pools of pathway components has been shown to regulate the function of both Arm and TCF proteins thus enabling them to assume multiple roles during development. Since both proteins are also direct binding partners of the nematode Nemo homolog, this interpretation of the phenotypes induced by loss of nmo must be taken within such a context (Verheyen, 2001 and references therein).

Much has been written about the relationships between signaling pathways during imaginal disc development and patterning. Only recently has insight been gained into the cross-regulation that occurs during the differentiation of vein and intervein fates during pupal stages. It is clear that complex regulatory interactions occur between the Egfr and Dpp vein promoting pathways and that Notch also plays an inhibitory role in Egfr signaling. Wingless signaling has to be added to the equation. It appears that Wg activity promotes veins and that nmo can antagonize this role. Interestingly, during embryogenesis Egfr and Wg signaling act antagonistically in the patterning of the denticle belts of the ventral epidermis, yet in pupal wings, initially both act to promote vein fates (Verheyen, 2001).

Some insight into the upstream regulation of Nemo has come from studies in worms and mice that suggest that the TGF-ß activated kinase (Tak-1) activates Nemo homologs in response to TGF-ß signaling (Meneghini, 1999; Ishitani, 1999; Shin, 1999). The finding that nmo and dpp mutations interact synergistically supports a role for Dpp in regulating Nemo activity in Drosophila, possibly through Tak-1. In addition, the similarity in phenotypes between overexpressed Nemo and mutations in the TGFß-like gene gbb-60A support a model whereby nmo may be differentially regulated by multiple TGFß pathways. Interestingly, another target of Tak-1, p38b, has also been shown to be involved in Dpp signaling in Drosophila and to play a role in wing vein specification (Verheyen, 2001 and references therein).

It is concluded that defects in nmo and genetic interactions have been observed that suggest the gene plays a role in two Wnt dependent processes in Drosophila: those controlling embryonic epidermal patterning and planar cell polarity. This is consistent with studies in C. elegans showing that the nmo homolog lit-1 displays a number of cell polarity defects, and acts in two separate Wnt signaling pathways (Rocheleau, 1999). Nemo homologs have been found to act at an intersection of TGFß and Wnt signaling. These studies suggest that Nemo interacts with other major developmentally pathways such as Notch and Egfr during eye and wing development. Current models of developmental patterning propose that signaling pathways are integrated at multiple levels and that this cross-talk is required for fine tuning of positional information. If these data are interpreted within this context, then Nemo may represent an important developmentally-regulated switch that allows precise spatial and temporal regulation of pattern formation (Verheyen, 2001).


The cDNAs were classified into two differentially spliced forms, type I and type II; they are alike in the 5' region but different in the 3' (Choi, 1994).

cDNA clone length - 2277 bases (clone C5-1)

Bases in 5' UTR - 573

Exons - 10

Bases in 3' UTR - 398 (clone C5-1)


Amino Acids - Type I and type II cDNAs have open reading frames that can encode polypeptides of 477 and 433 aa, respectively (Choi, 1994).

Structural Domains
The nmo gene sequence encodes a serine/threonine protein kinase homolog. The kinase domain of nmo is most similar to a family of MAP kinases, including murine ERK1 and yeast CDC28/cdc2-related genes with 47%-41% identity over a region of about 300 amino acids. The similarity of the nmo-encoded protein to MAP ikinases is limited to the catalytic domains (regions I through XI), so Nmo may not be functionally related to the MAP kinase family. Nmo shows differences from the ERK type: ERK protein kinases have two conserved residues, threonine and tyrosine, in subdomain VIII, a common autophosphorylation region; phosphorylation of both threonine (amino acid residue 200) and tyrosine (residue 202) is critical to their function. Nmo has the necessary threonine, but lacks the critical tyrosine (Choi, 1994).

nemo: Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

date revised: 4 August 2001

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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