InteractiveFly: GeneBrief

fussel: Biological Overview | References

Over the years Ski and Sno have been found to be involved in cancer progression e.g. in oesophageal squamous cell carcinoma, melanoma, oestrogen receptor-positive breast carcinoma, colorectal carcinoma, and leukaemia. Often, their prooncogenic features have been linked to their ability of inhibiting the anti-proliferative action of TGF-β signalling. Recently, not only pro-oncogenic but also anti-oncogenic functions of Ski/Sno proteins have been revealed. Besides Ski and Sno, which are ubiquitously expressed other members of Ski/Sno proteins exist which show highly specific neuronal expression, the SKI Family Transcriptional Corepressors (Skor). Among others Skor1 and Skor2 are involved in the development of Purkinje neurons and a mutation of Skor1 has been found to be associated with restless legs syndrome. But neither Skor1 nor Skor2 have been reported to be involved in cancer progression. Using overexpression studies in the Drosophila eye imaginal disc, this study analysed if the Drosophila Skor homologue fuss has retained the potential to inhibit differentiation and induce increased proliferation. fuss expressed in cells posterior to the morphogenetic furrow, impairs photoreceptor axon pathfinding and inhibits differentiation of accessory cells. However, if its expression is induced prior to eye differentiation, fuss might inhibit the differentiating function of Dpp signalling and might maintain proliferative action of Wg signalling, which is reminiscent of the Ski/Sno protein function in cancer (Rass, 2022).

Negative regulators of the TGF-β signalling pathway are inhibitory Smads (I-Smads), Smurfs and the Ski/Sno protein family. Proteins of the latter group possess two structural domains: the Ski/Sno homology domain and the SMAD4-binding domain. With the help of these domains, Ski/Sno proteins can interact, among others, with R-Smads, N-CoR, Sin3a, SMAD4 and the histone deacetylase HDAC1 and this complex leads to transcriptional repression of target genes. By their expression domains, Ski/Sno proteins can be further subdivided into ubiquitously expressed genes (human Ski and Sno), and mainly neuronally expressed genes, the SKI Family Transcriptional Corepressors (Skor1 and Skor2). The Ski/Sno proteins fulfil a wide range of different physiological functions such as axonal morphogenesis, Purkinje cell development, myogenesis and mammary gland alveogenesis (Rass, 2022).

However, the Ski/Sno proteins were not discovered by their physiological functions but via the transforming capability of the viral ski (v-ski) homologue found in the Sloan-Kettering virus. The first evidence that Ski/Sno proteins possess oncogenic capabilities came from overexpression experiments, where it was shown that the truncation of v-ski is not responsible for the transformation of chicken embryo fibroblasts, but that overexpression of v-ski, Ski or Sno is sufficient for this transformation. Despite this background, their role in carcinogenesis is still not fully understood, if not even contradictory at times. Ski and Sno have been found to be upregulated in different types of cancer e.g. oesophagus squamous cell carcinoma, melanoma, and colorectal cancer. Further evidence for a pro-oncogenic role was found in downregulation analyses of Sno or Ski. This downregulation resulted in decreased tumour growth in breast cancer cells and pancreatic cancer cells. But as stated before, there is some objection that Ski and Sno function purely as oncogenes. Mice, which were heterozygous mutant for Ski or Sno, showed an increased level of tumour induction after carcinogen treatment. In metastatic non-small cell lung cancer, Ski expression is significantly reduced, whereas increased expression of Ski in these cells reduced the invasiveness inhibiting epithelial-mesenchymal transition. Therefore, this could reflect that the outcome of Ski or Sno expression in cancer cells is dependent on the cell type or the actual status of the cancer cells and cancer cells often exploit Ski or Sno to inhibit the anti-proliferative effects of TGF-β signalling. Whereas Ski or Sno have been found to be involved in a lot of different cancer types, there is sparse evidence for deregulation of Skor proteins in cancer cells. Endogenously, Skor proteins have been linked to neurodevelopmental processes. After Skor1 overexpression, genes involved in axonal guidance or post-synapse assembly were differentially expressed. Skor2 is important for cerebellar Purkinje cell differentiation as in Skor2 knockout mice dendrite formation of Purkinje cells was impaired. Pathophysiologically, Skor1 has mainly been linked to restless leg syndrome and localized scleroderma (Rass, 2022).

In Drosophila melanogaster, only one homologue of Ski and Sno, which is designated Snoo, and one homologue of Skor1 and Skor2, which is designated Fuss, exist. It has been recently shown that fuss is interacting with SMAD4 and HDAC1. In overexpression assays, fuss can inhibit Dpp signalling and endogenously, the Fuss/HDAC1 complex is required for bitter gustatory neuron differentiation and fuss mutant flies pause more often during walking. However, this study was interested whether the Skor/Fuss proteins retained their ability to inhibit differentiation and induce increased proliferation. For this purpose, fuss was overexpressed in differentiating cells of the eye imaginal disc, an excellent model tissue to study regulatory gene function in the context of carcinogenesis. This overexpression impaired photoreceptor axon guidance and inhibited the differentiation of accessory cells such as cone cells and primary pigment cells, which are all transformed into a basal pigment cell type. In a second approach, fuss overexpressing clones were generated early during development in the eye imaginal discs, when cells are still proliferating. This resulted in vast outgrowths of undifferentiated tissue of the eye imaginal disc because fuss overexpression most likely inhibited Dpp-signalling, a member of the TGF-β superfamily. This work shows that fuss retained the ability of Ski/Sno proteins to inhibit the antiproliferative effects of TGF-β signalling by analogous inhibition of Dpp-signalling, allowing proliferation to be sustained (Rass, 2022).

The overexpression of fuss posterior to the morphogenetic furrow with the GMR-Gal4 driver line resulted in a nearly complete loss of all cell types in the adult eye. During development, photoreceptor axons were not able to target the appropriate layers of the optic lobe anymore and cone cells, primary pigment cells and bristle cells were transformed into a basal pigment cell fate. This transformation was caused by the inhibition of sv expression, which is crucial for accessory cell differentiation. Additionally, increased apoptosis during pupal development lead to the removal of photoreceptors and lastly adult eyes only consisted of cells containing pigment granules. This lack of differentiation cannot be explained by the Dpp inhibiting role fuss exerts, when overexpressed, because inhibiting the Dpp signaling pathway via knockdown of Tkv or Med had no effect. Photoreceptor axon guidance is impaired, if Dpp signaling is disrupted in photoreceptors by the expression of the inhibitory Smad Dad. Thus, the observed photoreceptor axon guidance phenotype, when fuss is overexpressed with GMR, could indeed be a result of Dpp signaling inhibition. However, the loss of nearly all eye cell types is due to other effects (e.g. downregulation of sv and apoptosis) than Dpp signaling repression alone, because loss of Dpp signaling behind the morphogenetic furrow only results in mild patterning defects of the pupal retina. Nonetheless, the inhibition of cell differentiation has already been shown in other cancer models e.g., when two copies of the constitutive active form of the receptor tyrosine kinase dRETMEN2B are expressed with the GMR-Gal4 line, pupal retinas are devoid of any distinguishable cell types. This phenotype is indistinguishable from the phenotype of the pupal retinas generated by the overexpression of fuss via GMR-Gal4. In a screen for novel oncogenes from breast cancer patients, human transgenes have been overexpressed with the GMR-Gal4 driver line. Overexpression of human RPS12, a subunit of the small ribosomal subunit, whose expression is increased in various cancer types, leads also to a glazed eye phenotype. Therefore, different oncogenes can result in different outcomes when expressed with the GMR-Gal4 driver line and are not always leading to massive tissue overgrowth like the Yorkie overexpression. Most importantly, with this approach to overexpress fuss in cells which already were destined for acquiring a cell fate and have left the cell cycle, it was not possible to induce increased proliferation anymore, but could prevent cell differentiation (Rass, 2022).

Consequently, a more pluripotent cell type in the eye imaginal disc was used and fuss overexpressing clones were induced prior to the formation of the morphogenetic furrow. These results allowed the assumption that in this context, fuss overexpressing clones do not react to the antiproliferative effects of the Dpp morphogen anymore. Instead, wg expression and thus, proliferation promotion might be maintained. This leads to outgrowths of clonal tissue from the eye imaginal disc of third instar larvae, which showed an increased number of mitotic events. If these flies survived to adulthood, undifferentiated, extra tissue was visible in the complex eye (Rass, 2022).

An analogous mechanism can be observed in tumors which overexpress Ski or Sno. The TGF-β signaling pathway is also anti-proliferative, but this action is inhibited by the increased presence of Ski/Sno proteins. Therefore, the molecular mode of action is similar to the human Ski/Sno proteins. The function of Ski and Sno is highly context dependent, as they can fulfill an anti-oncogenic or pro-oncogenic role depending on the cancer type or status of the cancer. This was also observed with fuss overexpressing clones. Only when induced 48h after egg laying, was additional tissue found in late third instar larvae and only in eye imaginal discs, because there, Dpp counteracts the proliferative effects of Wg signaling. When fuss is overexpressed in the wing disc or after induction of the morphogenetic furrow differentiation is inhibited, this results in a wing with truncated veins or in a smooth eye surface. This is also underlined by RNAseq data from eye and wing imaginal discs, where fuss was overexpressed with the GMR-Gal4 and Nub-Gal4 driver line, respectively. In the eye dataset, wg expression in eye imaginal discs is not significantly different from control eye discs, whereas wg expression in fuss overexpression wing discs is significantly reduced in contrast to control wing discs (Rass, 2022).

Thus, this study showed that the Skor protein fussin Drosophila melanogaster still retained the function of the Ski/Sno proteins by inhibiting differentiation but inducing hyperproliferation. But the hallmarks of real tumorigenesis are lacking, because at some point during pupal development, proliferation stops, and these cells become protruding head tissue as it could be observed in complex eyes of surviving flies. Furthermore, there was no evidence of an epithelial-mesenchymal transition because fuss overexpressing clones maintained their epithelial fate. It will be of high interest if future studies can find similar results in overexpression studies for the vertebrate Skor proteins or detect increased expression of these proteins in specific cancer types (Rass, 2022).

Adult Movement Defects Associated with a CORL Mutation in Drosophila Display Behavioral Plasticity

The CORL family of CNS-specific proteins share a Smad-binding region with mammalian SnoN and c-Ski protooncogenes. In this family Drosophila CORL has two mouse and two human relatives. Roles for the mouse and human CORL proteins are largely unknown. Based on genome-wide association studies linking the human CORL proteins Fussel15 and Fussel18 with ataxia, this study tested the hypothesis that dCORL mutations will cause adult movement disorders. For initial tests, side by side studies were conducted of adults with the small deletion Df(4)dCORL and eight control strains. Deletion mutants exhibit three types of behavioral plasticity. First, significant climbing defects attributable to loss of dCORL are eliminated by age. Second, significant phototaxis defects due to loss of dCORL are partially ameliorated by age and are not due to faulty photoreceptors. Third, Df(4)dCORL males raised in groups have a lower courtship index than males raised as singles though this defect is not due to loss of dCORL. Subsequent tests showed that the climbing and phototaxis defects were phenocpied by dCORL(21B) and dCORL(23C) two CRISPR generated mutations. Overall, the finding that adult movement defects due to loss of dCORL are subject to age-dependent plasticity suggests new hypotheses for CORL functions in flies and mammals (Dimitriadou, 2020).

How do the data impact the hypothesis that the association of Fussel15 and Fussel18 with ataxia will be conserved and visible in dCORL adult mutants as defects in movement-based behaviors? First, the absence of a larval crawling phenotype for Df(4)dCORL was impressive. Most ataxias are not pediatric and thus the absence of a developmental component to the Df(4)dCORL adult climbing defect is consistent with the hypothesis. Second, the absence of any courtship defect attributable to dCORL in Df(4)dCORL adult males is not inconsistent with the hypothesis. The complex nature of courtship that includes modalities that are not movement-based such as olfaction and singing provide other potetial causes for this defect. In addition, sphinx and Glu-RA that lie within Df(4)dCORL have already been implicated in courtship. Third, the phototaxis defect attributable to dCORL is not caused by a photoreceptor defect and phototaxis is regulated by the same neurotransmitters that regulate movement - dopamine and octopamine. This result is also consistent with the hypothesis (Dimitriadou, 2020).

Direct evidence supports the hypothesis: identification of adult climbing defects in two independenly generated dCORL mutants and their age-dependent plasticity. The defects in climbing are attributable to loss of dCORL due to assays with dCORL^23C and dCORL^21B and supported by the absence of climbing defects in six control strains with mutations in the region. It is concluded that the hypothesis is true, dCORL mutations cause deficits in adult movement-based behaviors (Dimitriadou, 2020).

Given this, are there implications for Fussel15 and Fussel18 in these data? First to clarify, data from transgene assays in flies showed that mCORL2/SKOR2 and dCORL are both capable of fully rescuing an endogenous function in Df(4)dCORL larvae while mCORL1/SKOR1 cannot. This fly transgene data indicates that movement defects in dCORL mutants are most relevant to Fussel18/SKOR2 associated ataxias. Second, it is important to consider the observations of 'attention-deficit' as the cause of the climbing and phototaxis defects. It is noted that initiation of both behaviors appeared normal but then Df(4)dCORL flies would wander randomly. Third, the age-dependent increases in climbing speed that erase the initial deficits in average score and average best score in males and females of all dCORL mutant genotypes could reflect 'hyperactivation' of the locomotor circuitry underlying walking. The last piece to consider is that the inbred spontaneously hypertensive rat is the most widely studied rodent model of Attention Deficit Hyperactivity Disorder (ADHD) and that these rats display both motor impairments and loss of Purkinje cells (Bruchhage, 2018). In the absence of a genetic model for ADHD, it is proposed that the mCORL2/SKOR2 knockout mouse is a candidate. Further, the data suggest that polymorphisms in Fussel18/SKOR2 be examined for association with ADHD in humans (Dimitriadou, 2020).

In an initial attempt to identify the mechanism behind the movement defects in dCORL mutants flies were pre-fed with yohimbine, an antagonist of tyramine receptors. In previous reports, yohimbine feeding rescued adult climbing and flight maintenance defects in a mutant characterized by low octopamine and high tyramine levels. Tyramine is a metabolic precursor and functional antagonist of octopamine, the fly counterpart to norepinephrine in mammals. In each of the three climbing assays, the combination of age and yohimbine improved the scores of Df(4)dCORL flies over age alone. How dCORL might interact with a tyramine/octopamine locomotion circuit awaits further experimentation (Dimitriadou, 2020).

Thinking broadly, a new hypothesis derived from the preliminary yohimbine data are that treatment of Fussel18/SKOR2 associated ataxia with a norepinephrine antagonist would be therapeutic. A first step would be to test this hypothesis in mCORL2/SKOR2 knockout mice with an ungainly gait from birth to adulthood. Rescue of the gait defect, analogous to the rescue of the climbing defects in dCORL mutants by age and yohimbine, is predicted if the hypothesis is true (Dimitriadou, 2020).

Overall, the data shows that Df(4)dCORL mutants exhibit three types of behavioral plasticity. First, significant climbing defects attributable to loss of dCORL are eliminated by age. Second, significant phototaxis defects due to loss of dCORL are partially ameliorated by age and are not due to faulty photoreceptors. Third, Df(4)dCORL males raised in groups have a lower courtship index than males raised as singles though this defect is not due to loss of dCORL. Subsequent tests showed that the climbing and phototaxis defects were phenocopied by dCORL^21B and dCORL^23C two CRISPR generated mutations. Overall, the finding that adult movement defects due to loss of dCORL are subject to age-dependent plasticity suggests new hypotheses for CORL functions in flies and mammals (Dimitriadou, 2020).

The Drosophila fussel gene is required for bitter gustatory neuron differentiation acting within an Rpd3 dependent chromatin modifying complex

Members of the Ski/Sno protein family are classified as proto-oncogenes and act as negative regulators of the TGF-β/BMP-pathways. A newly identified member of this protein family is fussel (fuss), the Drosophila homologue of the human functional Smad suppressing elements (fussel-15 and fussel-18). fuss interacts with SMAD4 and overexpression leads to a strong inhibition of Dpp signaling. fuss is a predominantly nuclear, postmitotic protein, mainly expressed in interneurons and fuss mutants are fully viable without any obvious developmental phenotype. fuss expression was characterized in the adult proboscis, and by using food choice assays it was possible to show that fuss mutants display defects in detecting bitter compounds. This correlated with a reduction of gustatory receptor gene expression providing a molecular link to the behavioral phenotype. In addition, fuss interacts with Rpd3, and downregulation of rpd3 in gustatory neurons phenocopies the loss of fuss expression. Surprisingly, there is no colocalization of fuss with phosphorylated Mad in the larval central nervous system, excluding a direct involvement of fuss in Dpp/BMP signaling. This work reveals fuss as a pivotal element for the proper differentiation of bitter gustatory neurons acting within a chromatin modifying complex (Rass, 2019).

The molecular and cellular functions of the fuss genes, which are members of the Ski/Sno protein family, are still poorly understood. The fact that Drosophila contains only one single fuss gene offers a great opportunity for a thorough analysis. However, this has been restrained due to its location on the 4^th chromosome, where only limited genetic tools were available. As a consequence, previous reports have been focusing on the analysis of either overexpression studies or by using a multi-gene deficiency with contradictory results. In the meantime, more recent methodological advances like the CRISPR/Cas9 genome editing and the MiMIC gene trap technique have expanded the Drosophila genetic toolbox and provided an appropriate genetic environment allowing a thorough and in-depth study of such genes. The availability of the fuss^Mi13731 fly line, which is a gene trap of fuss, allowed study of the expression pattern of Fuss. This line perfectly matches Fuss-antibody stainings and was used to create a Gal4 line via Recombinase-mediated cassette exchange (RMCE). A second independent mutant fuss allele, fuss^delDS was created by CRISPR/Cas9 editing by deletion of the main functional protein domains. Although fuss^Mi13731 and fuss^delDS alleles are generated by different genetic approaches they share the same phenotypes, underlining that despite the complex genomic organization of fuss the observed phenotypes are due to the loss of fuss. Surprisingly, fuss mutant flies are fully viable and do neither show developmental lethality or reduced lifespans nor any other apparent phenotypes (Rass, 2019).

By means of these new tools, it was possible show that Fuss is expressed postmitotically in a small subset of neurons. All Fuss neurons in the CNS are interneurons, but they express different cell fate markers, suggesting that they represent a rather diverse group of neurons. These results were confirmed molecularly by a targeted DamID experiment, which, in addition, indicated a highly specific expression of gustatory receptor genes and indeed, Fuss is expressed in one Gustatory Receptor Neuron (GRN) per sensillum. In S and I-type sensilla it is expressed in bitter GRNs and in L-type sensilla, which lack bitter GRNs, it is expressed in salt attracting GRNs. How the bitter GRNs react to the loss of Fuss was investigated, and interestingly, this leads to an impairment of bitter sensation. Remarkably, this phenotype is correlated with a downregulation of bitter gustatory receptors Gr33a, Gr66a and Gr93a and in some bitter GRNs of fuss mutant flies no Gr33a expression can be observed anymore. The expression of Fuss in sensory neurons during development, and the adult phenotype, suggest that Fuss is needed for the proper maturation of these neurons and therefore is essential for bitter GRN differentiation. As there is a possibility, that the bitter sensation phenotype might be due to some higher order interneurons within the CNS, a specific UAS-t::gRNA-fuss^4x line was generated to be able to perform cell type specific gene knockouts. Indeed, using an independent driver line (Poxn-Gal4-13-1) expressed in all GRNs, faithfully reproduced this phenotype indicating a direct association of bitter sensation and GRN defects. In fuss mutant flies morphology of bitter GRNs was not altered and cell number was just slightly changed compared to controls, while Gr33a expression was completely lost in 40% of all bitter GRNs and Gr66a expression was reduced in all GRNs, but was never completely absent from a bitter GRN. Therefore, in fuss mutant flies bitter GRNs are correctly specified but the terminal differentiation of this neurons is disturbed, which ultimately results in impaired bitter taste sensation. This is comparable to Fuss neurons in the larval and adult CNS, where loss of Fuss expression also did not have an impact on axonal projections or cell numbers and thus not on initial specification of these neurons. This supports the idea, that Fuss is required for fine tuning individual subgroups of neurons during development, a phenotype, which resembles loss of Skor2 in mice, where it is dispensable for initial Purkinje cell fate specification but is required for proper differentiation and maturation of Purkinje cells (Nakatani, 2014). It is very likely that other genes will also be affected by the loss of Fuss, and the reduction of these gustatory receptors could lead to a cumulative effect, as it has been shown that they act in heteromultimers where a multimeric receptor consists of at least Gr66a, Gr33a and Gr93a, which are all required for caffeine sensation. Whereas over the years many studies have dissected the function of single gustatory receptors, the complexes they establish, and genes which are involved in more common topics like sensory neuron formation, less is known about the differentiation and specification of subsets of GRNs. To find further genes involved in differentiation of bitter GRNs and to clarify the molecular consequences of the fuss mutation in bitter GRNs, transcriptional profiling experiments will be conducted specifically in Fuss positive GRNs (Rass, 2019).

Using the targeted DamID (TaDa) method, it was of interest to see if this method is sensitive enough to pick up significant differences between fuss mutant and wildtype flies. This was indeed the case for GR66a. However, in general, the performed TaDa experiments showed only slight differences between mutant and control flies. This could be a consequence of Fuss being expressed in heterogenic neuronal clusters. This study showed, that Fuss interacts with Rpd3, a histone deacetylase, and therefore, a chromatin modifier, which is preferentially associated with inhibitory gene regulating complexes. This could be a common mechanism for Fuss in all Fuss expressing neurons. However, different neuronal populations have different open and closed chromatin and probably the Fuss/Rpd3 complex regulates different genes in different neuronal populations, which could lead to the masking of differential gene expression by individual neuronal cell groups. Additionally, although the TaDa technique functions very well to generate transcriptional profiles without cell isolation, data is nondirectional and at GATC fragment resolution, which decreases overall resolution. To overcome these limitations experiments are on the way to unravel the function of specific neuronal clusters as well as the function of fuss in these neuronal clusters and to specifically profile transcription of these clusters and changes upon loss of fuss (Rass, 2019).

A careful analysis with the newly generated antibodies shows that there is no expression of Fuss in larval or adult Kenyon cells as has been postulated recently. To unequivocally show that there is no requirement for Fuss in mushroom body development, neither autonomously nor non-autonomously, Fuss expressing neurons were ablated using a fuss-GAL4 line driving Reaper. Again, these flies, even without any fuss expressing cells, are fully viable and do not show mushroom body defects. Lastly, no evidence was found of Fuss being expressed in insulin producing neurons by antibody staining or DamID experiments as shown recently. These discrepancies are most likely explained by the use of the specific knockout line fuss^delDS, and the gene trap line fuss^Mi13731 in the current case, whereas a 40 kb genomic deletion Df(4)dCORL was used in previous studies. This deletion covered the fuss locus as well as two more protein coding genes, 4E-T and mGluR, and three noncoding RNA genes, CR45201, CR44030 and sphinx. Any of these, or a combination of them, could be responsible for premature lethality or mushroom body defects. One additional possible explanation for their mushroom body defects in the deletion is an inappropriate fusion of a new transcriptional start site or enhancer region from the mGluR upstream to the toy gene creating a weak overexpression phenotype of toy in mushroom bodies, a phenotype, which has been described already. Indeed, very recently Tran (2018) described a slight overexpression of Toy in their deficiency allele Df(4)dCORL (Rass, 2019).

It has been shown that Ski/Sno proto-oncogenes have an inhibitory effect on TGF-β or BMP signaling in overexpression assays. This is often associated with the ability of Ski/Sno proteins to inhibit the antiproliferative effects of TGF-β signaling in cancer and to promote their progression. However, in an endogenous situation, Fuss is not expressed in cells, where the BMP/Dpp signaling pathway is active. This is displayed by the absence of the motoneuron marker pMad in Fuss neurons. Later in adulthood, Mad itself is also not specifically enriched in Fuss expressing neurons according to the TaDa dataset, clearly pointing against a function in BMP signalling. An interaction between Fuss and Smox was tested in CoIP assays. However, the possibility cannot be ruled out that the phosphorylated form of Smox is interacting with Fuss or the Fuss/Med complex. But since both phosphorylated Smox and Fuss interact with Medea, it would also be possible to get an artificial interaction. At least according to the TaDa dataset, Smox is expressed in Fuss neurons. Unfortunately, there is currently no good marker available to test for an activated TGF-β signaling pathway in Drosophila cells, like an antibody against phosphorylated Smox. What might be the main molecular mechanism for Fuss? Although the Ski/Sno/Dac homology domain and the SMAD4 binding domain in Ski have DNA binding character, they mainly have been shown to be involved in protein-protein interactions (Nyman, 2010). Furthermore, Ski/Sno proteins do not possess an intrinsic catalytic activity, they rather act as recruiting proteins (Deheuninck, 2009). In agreement, this study could show that this is also the case for Fuss. Not only that Fuss binds to Medea, which is a DNA binding protein and therefore mediates the DNA binding, Fuss also interacts with Rpd3, a histone deacetylase. Thus, the Med/Fuss/Rpd3 complex is involved in chromatin silencing and plays a key role in terminal differentiation. Interestingly, the loss of bitter sensation and downregulation of bitter GRs could also be phenocopied by a knockdown of rpd3 in Fuss expressing gustatory neurons. One current hypothesis of Fuss/Rpd3 function in GRNs is, that this protein complex is inhibiting a repressor of GR genes and in the absence of either fuss or rpd3, the complex is inactivated and this repressor will inhibit bitter GR genes (Rass, 2019).

For Ski and Sno, the transcriptional repressor complexes have been reasonably well characterized (Tecalco-Cruz, 2018), but for the Fuss-type proteins, very little is known about their complexes. It would be highly interesting if Fuss proteins act through repressor complexes identical to the complexes of Ski or Sno or a rather unique one. The most exciting question to solve regarding protein interaction will be, if the Fuss/Rpd3 complex plays a role in TGF-β signalling, or if in contrast to its mammalian homologues, it is not only acting BMP independent, but also independent from the TGF-β signalling cascade. Besides identifying further protein-protein interactions and investigating DNA-protein interactions more precisely, it will be very important to describe the exact function of the Fuss/Rpd3 complex. In mammals, Skor2 is thought to activate Sonic Hedgehog expression in Purkinje cells from direct binding to the Sonic Hedgehog promotor and this might be achieved by inhibition of the BMP pathway or by cooperation with the RORalpha pathway, a nuclear orphan receptor. In contrast to that, Skor1 interacts with Lbx1, a homologue of the ladybird early or ladybird late in Drosophila, and acts as a transcriptional corepressor of Lbx1 target genes. The TaDa datasets strongly point towards another function for Fuss in Drosophila, as neither hedgehog nor the homologues of Lbx1, ladybird late and ladybird early, are enriched in Fuss expressing cells. Therefore, identifying target genes, interacting proteins, binding motifs of the Fuss complex and subsequent comparison with established models for other transcription factor complexes will elucidate the role of this complex in cell fate determination (Rass, 2019).

CORL expression in the Drosophila central nervous system is regulated by stage specific interactions of intertwined activators and repressors

CORL proteins (SKOR in mice and Fussel in humans) are a subfamily of central nervous system (CNS) specific proteins related to Sno/Ski oncogenes. Their developmental and homeostatic roles are largely unknown. Previous work has shown that Drosophila CORL (dCORL; fussel in Flybase) functions between the Activin receptor Baboon and Ecdysone Receptor-B1 (EcR-B1) activation in mushroom body neurons of third instar larval brains. To better understand dCORL regulation and function a series of reporter genes was generated. this study examined the embryonic and larval CNS and found that dCORL is regulated by stage specific interactions between intertwined activators and repressors spanning numerous reporters. The reporter AH.lacZ, which contains sequences 7-11kb upstream of dCORL exon1, reflects dCORL brain expression at all stages. Surprisingly, AH.lacZ is not present in EcR-B1 expressing mushroom body neurons. In larvae AH.lacZ is coexpressed with Elav and the transcription factor Drifter as well as in dILP2 insulin producing cells of the pars intercerebralis. The presence of dCORL in insulin producing cells suggests that dCORL functions non-autonomously in the regulation of EcR-B1 mushroom body activation via the modulation of insulin signaling. Overall, the high level of sequence conservation seen in all CORL/SKOR/Fussel family members and their common CNS-specificity suggest that similarly complex regulation and a potential function in insulin signaling are associated with SKOR/Fussel proteins in mammals (Tran, 2018).

Transgenic analyses in Drosophila reveal that mCORL1 is functionally distinct from mCORL2 and dCORL

Uncovering how new members of multigene families acquire new functions is an important topic in evolutionary and developmental genetics. CORL proteins (SKOR in mice, Fussel in humans and fussel in Drosophila) are a family of CNS specific proteins related to mammalian Sno/Ski oncogenes. Drosophila CORL (dCORL/Fussel) participates in TGF-beta and insulin signaling during development and in adult homeostasis but roles for the two mouse CORL proteins (mCORL) are essentially unknown. A series of studies were conducted to test the hypothesis based on previous results that mCORL1 is more similar to dCORL than mCORL2. Neither an updated alignment nor ectopic expression in adult wings were able to distinguish mCORL1 or mCORL2 from dCORL. Transgene experiments employing a dCORL endogenous function in mushroom body neurons showed that mCORL1 is distinct from mCORL2 and dCORL. mCORL1 and mCORL2 are also distinct in biochemical assays of Smad-binding and BMP signaling. Taken together, the data suggests testable new hypotheses for mCORL2 function in mammalian TGF-beta and insulin signaling based on known roles for dCORL. Overall, the study reiterates the value of transgenic methods in Drosophila to provide new information on multigene family evolution and the function of family members in other species (Stinchfield, 2019).

fussel (fuss)--a negative regulator of BMP signaling in Drosophila melanogaster

The TGF-β/BMP signaling cascades control a wide range of developmental and physiological functions in vertebrates and invertebrates. In Drosophila melanogaster, members of this pathway can be divided into a Bone Morphogenic Protein (BMP) and an Activin-β (Act-β) branch, where Decapentaplegic (Dpp), a member of the BMP family has been most intensively studied. They differ in ligands, receptors and transmitting proteins, but also share some components, such as the Co-Smad Medea (Med). The essential role of Med is to form a complex with one of the two activating Smads, Mothers against decapentaplegic (Mad) or dSmad, and to translocate together to the nucleus where they can function as transcriptional regulators of downstream target genes. This signaling cascade underlies different mechanisms of negative regulation, which can be exerted by inhibitory Smads, such as Daughters against decapentaplegic (Dad), but also by the Ski-Sno family. This work identified and functionally analyzed a new member of the Ski/Sno-family, fussel (fuss, annotated as CG11093, currently termed CORL (Now termed Fussel), the Drosophila homolog of the human functional suppressing element 15 (fussel-15). fuss codes for two differentially spliced transcripts with a neuronal expression pattern. The proteins are characterized by a Ski-Sno and a SAND homology domain. Overexpression studies and genetic interaction experiments clearly reveal an interaction of fuss with members of the BMP pathway, leading to a strong repression of BMP-signaling. The protein interacts directly with Medea and seems to reprogram the Smad pathway through its influence upon the formation of functional Mad/Medea complexes. This leads among others to a repression of downstream target genes of the Dpp pathway, such as optomotor blind (omb). Taken together shows that fuss exerts a pivotal role as an antagonist of BMP signaling in Drosophila melanogaster (Fischer, 2012).

This report characterized a new gene in Drosophila melanogaster, fussel (fuss), an ortholog of the human functional smad suppressing element 15 (fussel-15), also known as SKOR1, Corl1 or LBXcor1. fussis characterized by a Ski- Sno and a SAND homology domain and can be classified as a proto-oncogene. The two transcripts, fussB and fussC, diverge in the N-terminus and represent two phylogenetically different versions of the gene: fussB is the original form of the CG11093 locus, whereas the fussC transcript is spliced differently due to the subsequent integration of a Tc1-2 transposon. The ubiquitous mis-expression of both forms is lethal in pupal stages. Its endogenous expression pattern during embryogenesis and larval development is neuronal, which is similar to the vertebrate genes fussel-15 and fussel-18 which also show a restricted pattern of expression mainly limited to neuronal tissue such as the developing murine cerebellum and the spinal cord (Fischer, 2012).

As Ski-Sno proteins are described to repress TGF-β signaling through their interaction with Smad proteins this study investigated if fuss is able to inhibit the TGF-β/BMP cascade, which is represented in Drosophila by BMP/Activin-β signaling. Use was made of the wing blade and the adult wing, an amenable and widely used tissue to analyse function and crosstalk of members within this signaling pathway. The ectopic expression of fuss in the wing affects the overall vein structure. While fussC produces defects in L2, L5 and p-cv and results in a loss of the campaniform sensilla on L3, the fussB isoform overexpression leads to a complete wing disruption or poorly unfolded extremities in the few female escapers which hatch. Together these phenotypes are highly reminiscent of loss of function phenotypes within the dpp pathway and provoked an investigation of if and how fuss is able to interact negatively within this cascade. While the contribution of BMP signaling to Drosophila development is enormous, including cell-fate specification, imaginal disk patterning or growth organization, the Activin-β branch has only been elucidated recently. It could be shown, that its components regulate neuronal wiring and proliferation, mushroom body remodeling and the morphogenesis of neurons in the adult. It was of interest to see if it was possible to decipher the pathway affected by fuss overexpression and examined the expression pattern of prominent read-outs, namely omb and sal for BMP signaling and EcR1B for the Activin-β cascade. The results show a clear reduction of omb and sal on a histological and also molecular level, which leads to the conclusion that fuss is indeed an inhibitor of the BMP pathway. Moreover, the coexpression of dpp-cascade activators like the typeI receptor saxophone or the Smads mad and medea with fuss results in a clear rescue of vein patterning and wing size. In the case of medea an almost complete rescue of the A9-Gal4;UAS-med wing phenotype was observed and a direct interaction of Fuss with Medea via its SMAD4 binding domain was postulated, as has been described for dSno or c-ski. In contrast to the BMP pathway, no effects could be observed on the expression of one of the main target genes of the Activin-β pathway, EcR1b. This result was further supported by a failure to detect genetic interactions of fuss with members of the Activin-β branch, which further supports a specific inhibitory function of fuss on the BMP pathway. To further strengthen the hypothesis of a specific interaction of Fuss with Medea, the direct interaction of these two proteins was identified by a yeast two hybrid experiment and confirmed by CoIP in Drosophila cell culture. Interestingly an interaction between Fuss and Mad in vitro, although the genetic interaction of Fuss with Mad revealed a partial rescue of the wing phenotype. The fact that pMad concentration in wing disks is not reduced in the presence of fuss clearly indicates, that fuss function is downstream of R-Smad activation. One possibility could be that Fuss is able to titrate out pMad/Med or possibly forms a trinary complex with pMad/Med affecting BMP target gene expression like omb (Fischer, 2012).

Is there a functional difference between dSno and Fuss, both belonging to the ski family? Although dSno exerts its effects also through Med, it is supposed to act as a BMP-to-Activin-β pathway switch, at least in brain development. By forming a complex with dSMAD2 it directs differentiation of neuroblasts towards proliferation, a role we can not ascertain for Fuss. However, a clear difference in vein patterning defects was observed comparing dSno and fuss overexpression using an identical driver line (A9- Gal4) further supporting an individual and different inhibitory effect of the closely related Ski/Sno/Fuss proteins (Fischer, 2012).

The interaction of Fuss with Med and the subsequent inhibitory effects on BMP signaling, led to the assumption, that the subcellular localization of Fuss might undergo changes during dpp activation transmitted by med overexpression. In general, the subcellular localization of homologous proteins such as Ski and SnoN is variable and depends on several conditions, such as morphological differentiation of cells or activity of proteins in normal versus tumor tissues; for example SnoN localization in nontumorigenic cells is preferentially cytoplasmic, while in tumor cells it is constitutively nuclear. GFP tagged fuss protein is predominantly localized in the cytoplasm, when it is overexpressed by itself. Here it might sequester Med and prevent its nuclear translocation in response to dpp signaling or another yet unidentified factor. However, overexpression of med together with fuss leads to a clear relocalization from the cytoplasm into the nucleus. fuss thereby antagonizes the BMP cascade, which is overstimulated by excessive med signaling leading to an almost complete rescue of the med overexpression wing phenotype (Fischer, 2012).

Interestingly, it was observed that pMad still translocates into the nucleus upon fuss overexpression. Considering the genetic interaction results, it is very likely that pMad/Med/Fuss enter the nucleus already as a trimer, which then might lead to a change in DNA binding or regulatory capabilities of the Smad complex. The results also need to be discussed in respect to recent data on the mammalian Fussel genes, in particular the isolation and characterization of a transposon induced mouse null allele of Fussel-18 (Skorl-2). Interestingly one prominent signaling phenotype in these mice is a strong repression of Sonic hedgehog (Shh) signaling. In particular it was shown that Fussel-18 is able to bind R-Smads and Co-Smads leading to a specific reduction of BMP- but not TGF-β-signaling. This nicely fits the genetic interaction results, which show that both Mad and Med interact with Fuss, although it was only possible to show physical protein interaction for Med and not for Mad. The Shh repression effect in the mouse mutant can be explained through a repressing function of BMP signaling on Shh, which in a wildtype background, is repressed by Fussel-18 itself, reestablishing Shh expression (Fischer, 2012).

The exact mechanism through which fuss exerts its endogenous function remains to be elucidated. As mentioned before, the protein exhibits a DHD motif, known to be responsible for DNA binding. Although a direct interaction with DNA could neither be shown for the human Ski-Sno proteins, nor for Drosophila SNO the possibility that fuss translocates to the nucleus (with or without Med) and binds itself to DNA cannot not rule out. Yet it is rather proposed an association of fuss with other protein partners, such as the corepressors Smrter or dSin3A or that it forms a complex with the histone deacetylase Rpd3, such an association of Fussel-18 with HDAC1 has been described in the mouse. It is also possible that fuss displaces coactivators, such as the Drosophila homolog of the p300/CBP complex called Nejire or stabilizes inactive SMAD complexes (Fischer, 2012).

Further investigations will elucidate the endogenous role of fuss during Drosophila development and the processes by which it antagonizes BMP signaling (Fischer, 2012).

Fussel-15, a novel Ski/Sno homolog protein, antagonizes BMP signaling. Mol Cell

The Ski family of nuclear oncoproteins represses transforming growth factor-beta (TGF-beta) signaling through inhibition of transcriptional activity of Smad proteins. This study identified a novel gene, fussel-15 (functional smad suppressing element on chromosome 15) with high homology to the recently discovered Fussel-18 protein. Both, Fussel-15 and Fussel-18, share important structural features, significant homology and similar genomic organization with the homolog Ski family members, Ski and SnoN. Unlike Ski and SnoN, which are ubiquitously expressed in human tissues, Fussel-15 expression, like Fussel-18, is much more restricted in its expression and is principally found in the nervous system of mouse and humans. Interestingly, Fussel-15 expression is even more restricted in adulthood to Purkinje cells of human cerebellum. In contrast to Fussel-18 that interacts with Smad 2, Smad3 and Smad4 and has an inhibitory activity on TGF-beta signaling, Fussel-15 interacts with Smad1, Smad2 and Smad3 molecules and suppresses mainly BMP signaling pathway but has only minor effects on TGF-beta signaling. This new protein expands the family of Ski/Sno proto-oncoproteins and represents a novel molecular regulator of BMP signaling.

Search PubMed for articles about Drosophila Fussel

Arndt, S., Poser, I., Moser, M. and Bosserhoff, A. K. (2007). Fussel-15, a novel Ski/Sno homolog protein, antagonizes BMP signaling. Mol Cell Neurosci 34(4): 603-611. PubMed ID: 17292623

Bruchhage, M. M. K., Bucci, M. P. and Becker, E. B. E. (2018). Cerebellar involvement in autism and ADHD. Handb Clin Neurol 155: 61-72. PubMed ID: 29891077

Deheuninck, J. and Luo, K. (2009). Ski and SnoN, potent negative regulators of TGF-beta signaling. Cell Res 19(1): 47-57. PubMed ID: 19114989

Dimitriadou, A., Chatzianastasi, N., Zacharaki, P. I., O'Connor, M., Goldsmith, S. L., O'Connor, M. B., Consoulas, C. and Newfeld, S. J. (2020). Adult Movement Defects Associated with a CORL Mutation in Drosophila Display Behavioral Plasticity. G3 (Bethesda). PubMed ID: 32161085

Fischer, S., et al. (2012). fussel (fuss)--a negative regulator of BMP signaling in Drosophila melanogaster. PLoS One 7(8): e42349. PubMed Citation: 22879948

Nakatani, T., Minaki, Y., Kumai, M., Nitta, C. and Ono, Y. (2014). The c-Ski family member and transcriptional regulator Corl2/Skor2 promotes early differentiation of cerebellar Purkinje cells. Dev Biol 388(1): 68-80. PubMed ID: 24491816

Nyman, T., Tresaugues, L., Welin, M., Lehtio, L., Flodin, S., Persson, C., Johansson, I., Hammarstrom, M. and Nordlund, P. (2010). The crystal structure of the Dachshund domain of human SnoN reveals flexibility in the putative protein interaction surface. PLoS One 5(9): e12907. PubMed ID: 20957027

Rass, M., Oestreich, S., Guetter, S., Fischer, S. and Schneuwly, S. (2019). The Drosophila fussel gene is required for bitter gustatory neuron differentiation acting within an Rpd3 dependent chromatin modifying complex. PLoS Genet 15(2): e1007940. PubMed ID: 30730884

Rass, M., Gizler, L., Bayersdorfer, F., Irlbeck, C., Schramm, M. and Schneuwly, S. (2022). The Drosophila functional Smad suppressing element fuss, a homologue of the human Skor genes, retains pro-oncogenic properties of the Ski/Sno family. PLoS One 17(1): e0262360. PubMed ID: 35030229

Stinchfield, M. J., Miyazawa, K. and Newfeld, S. J. (2019). Transgenic analyses in Drosophila reveal that mCORL1 is functionally distinct from mCORL2 and dCORL. G3 (Bethesda). PubMed ID: 31530634

Tecalco-Cruz, A. C., Rios-Lopez, D. G., Vazquez-Victorio, G., Rosales-Alvarez, R. E. and Macias-Silva, M. (2018). Transcriptional cofactors Ski and SnoN are major regulators of the TGF-beta/Smad signaling pathway in health and disease. Signal Transduct Target Ther 3: 15. PubMed ID: 29892481

Tran, N. L., Takaesu, N. T., Cornell, E. F. and Newfeld, S. J. (2018). CORL expression in the Drosophila central nervous system is regulated by stage specific interactions of intertwined activators and repressors. G3 (Bethesda). Pubmed ID: 29848623

date revised: 27 October 2022

Home page: The Interactive Fly © 2011 Thomas Brody, Ph.D.