InteractiveFly: GeneBrief

Ets at 21C: Biological Overview | References


Gene name - Ets at 21C

Synonyms -

Cytological map position - 21E2-21E2

Function - transcription factor

Keywords - stress-induced transcription factor - controls intestinal stem cell proliferation & enterocyte apoptosis through distinct sets of target genes - loss of function prevents epithelial aging - ectopic expression increases tumor size - expression is regulated by the JNK pathway - Capicua represses cell proliferation via direct targets including Ets21C

Symbol - Ets21C

FlyBase ID: FBgn0005660

Genetic map position - chr2L:547,016-552,923

NCBI classification - ETS: erythroblast transformation specific domain

Cellular location - nuclear



NCBI links: EntrezGene, Nucleotide, Protein
BIOLOGICAL OVERVIEW

Homeostatic renewal and stress-related tissue regeneration rely on stem cell activity, which drives the replacement of damaged cells to maintain tissue integrity and function. The Jun N-terminal kinase (JNK) signaling pathway has been established as a critical regulator of tissue homeostasis both in intestinal stem cells (ISCs) and mature enterocytes (ECs), while its chronic activation has been linked to tissue degeneration and aging. This study shows that JNK signaling requires the stress-inducible transcription factor Ets21c to promote tissue renewal in Drosophila. Ets21c controls ISC proliferation as well as EC apoptosis through distinct sets of target genes that orchestrate cellular behaviors via intrinsic and non-autonomous signaling mechanisms. While its loss appears dispensable for development and prevents epithelial aging, ISCs and ECs demand Ets21c function to mount cellular responses to oxidative stress. Ets21c thus emerges as a vital regulator of proliferative homeostasis in the midgut and a determinant of the adult healthspan (Mundorf, 2019).

The intestinal epithelium undergoes continuous homeostatic and acute, stress-induced cellular turnover to ensure tissue integrity and function throughout an organism's lifetime. The replacement of damaged and aberrant cells is fueled by somatic stem cells, whose proliferation is tightly controlled and coordinated with differentiation to satisfy tissue needs while preventing organ degeneration or tumor formation. In the Drosophila adult midgut, which is a functional equivalent of the mammalian small intestine, the intestinal stem cells (ISCs) divide asymmetrically to self-renew and generate two different cell types: the transient enteroblasts (EBs) and the enteroendocrine (EE) lineage-determined cells. Following several rounds of endoreplication, the EBs mature into the large, polyploid, and polarized enterocytes (ECs), which represent the major building blocks of the midgut epithelium. While primarily involved in nutrient resorption, the ECs also serve as a physical and chemical barrier protecting the organism against toxins, pathogens, oxidative stress, and mechanical damage. The runaway stem cell activity and loss of intestinal integrity due to chronic inflammation and increased stress load have been recognized as the prime underlying causes of aging-associated tissue decline and lifespan shortening (Biteau, 2008, Biteau, 2010, Guo, 2014). How stress signals are transduced and integrated with the homeostatic maintenance mechanisms at the cellular level and the organ level is only partially understood (Mundorf, 2019).

The evolutionarily conserved Jun N-terminal kinase (JNK) signaling is among the key pathways that govern regenerative responses to stress, infection, and damage in the intestine. Its chronic activation has been linked to the breakdown of epithelial integrity and accelerated aging. JNK signaling affects both ISCs and differentiated ECs. In the ECs, JNK confers stress tolerance and promotes epithelial turnover by triggering the apoptosis of damaged ECs and compensatory ISC proliferation. At the same time, cell-autonomous JNK activation in ISCs accelerates ISC mitosis in cooperation with the epidermal growth factor receptor (EGFR/Ras/ERK) signaling pathway, which provides the permissive signal for division. In contrast, JNK suppression prevents age-associated ISC hyperproliferation, accumulation of mis-differentiated cells, and epithelial dysplasia, resulting in lifespan extension. The canonical response to JNK signaling culminates in the activation of transcription factors that orchestrate gene expression. The basic leucine zipper (bZIP) transcription factors Fos (kayak) and Jun (jun-related antigen) are the best-characterized JNK pathway transcriptional effectors during development. In the adult intestine, however, the relation between JNK, Jun, and Fos is less clear. Deficiency for either Fos or Jun interferes with ISC survival, a response that is not observed upon JNK inhibition. In addition, the transcription factor Foxo has been shown to orchestrate adaptive metabolic responses downstream of JNK in ECs. However, Foxo overexpression does not drive epithelial renewal as JNK activation does. These data strongly suggest that other transcription factors may play a role in mediating the pleiotropic, adaptive JNK responses in the different cell types of the intestine (Mundorf, 2019).

The transcription factors of the E-twenty six (ETS) family are functionally conserved in all metazoans and are implicated in a plethora of processes, including cell-cycle control, differentiation, proliferation, apoptosis, and tissue remodeling. Several genome-wide sequencing approaches have determined that Drosophila ets21c, the ortholog of human proto-oncogenes FLI1 and ERG, is transcriptionally induced by infections, wounding, tumorigenesis, and aging (Blanco, 2010, Boutros, 2002, Broderick, 2014, Külshammer, 2015, Patterson, 2013). In the case of epithelial tumors and lipopolysaccharide treatment, ets21c upregulation required JNK activity (Boutros, 2002, Külshammer, 2015, Toggweiler, 2016), thus making Ets21c a plausible candidate to act as a JNK effector in the adult intestine (Mundorf, 2019).

This study shows that Ets21c acts as a critical and specific regulator of ISC and EC functions in the adult fly intestine downstream of JNK signaling and in response to oxidative stress and aging. Ets21c is necessary and sufficient to coordinate epithelial turnover by controlling ISC proliferation and the removal of ECs. By regulating specific sets of target genes, Ets21c orchestrates distinct cellular behaviors of midgut cells via intrinsic and non-autonomous signaling mechanisms. While dispensable for normal development, Ets21c functions as a vital determinant of stress tolerance and lifespan (Mundorf, 2019).

By targeted manipulation of Ets21c in progenitor cells, this study shows that Ets21c levels affect the rate of ISC proliferation intrinsically while their maintenance and survival remain unaltered. The reduced ERK activation in ISCs as a consequence of ets21c deficiency in the stress-free context could provide a mechanism for the observed decline in ISC mitotic capacity. This would be consistent with a described dependency of the JNK-induced ISC hyperproliferation on the EGFR/Ras/ERK signaling pathway. The precise mechanism by which Ets21c regulates ERK activity and ISC proliferation remains to be determined. However, the identification of pvf1 as a direct transcriptional target of Ets21c implies that Ets21c could modulate mitotic Pvr/Ras signaling in progenitors through the ISC-derived autocrine and EC-specific paracrine production of Pvf1 (Mundorf, 2019).

Differentiated ECs also require intrinsic Ets21c activity for proper function. EC-specific Ets21c activation drives epithelial turnover, which involves the apoptotic removal of mature ECs and ISC proliferation to renew the pool of differentiated cells. The Ets21c-mediated EC removal and compensatory ISC proliferation response could be suppressed by both co-expression of the pan-caspase inhibitor p35 or knockdown of the Ets21c target eip93F that controls Dcp1 activity. Neither upd3 nor pvf1 silencing in ECs interfered with Dcp1 activation, although both were indispensable for the non-autonomous induction of ISC proliferation by ets21c-expressing ECs. Based on these data, it is concluded that Ets21c orchestrates epithelial turnover by promoting EC apoptosis and stimulating compensatory ISC proliferation by apoptosis-dependent and -independent mechanisms exploiting intercellular signaling molecules such as Pvf1 growth factor and the chief stress-inducible cytokine Upd3. Apoptosis-dependent and -independent induction of ISC proliferation has also been demonstrated for JNK signaling in ECs. The cell death-independent mechanism would also explain why the ISC mitotic rate remains high in paraquat-exposed flies despite Dcp1 inactivation due to EC-specific ets21c silencing. In this respect, it is important to note that other transcription factors besides Ets21c have been shown to regulate Upd3 expression under stress conditions -- for example, in infected ECs or upon oncogene activation in imaginal discs (Mundorf, 2019).

Of note, increased JNK activity has been associated with age- and oxidative stress-related changes in the posterior midgut. Consistent with its role as a JNK-dependent transcriptional effector, Ets21c levels build up in response to paraquat and during aging. ISC/EB- and EC-specific TaDa profiling revealed that only a small fraction of the Ets21c-associated genes was actively transcribed, indicating that in the unstressed state, Ets21c contributes to the fine-tuning of gene expression that supports the steady-state epithelial replenishment. Its seemingly 'unproductive' binding to DNA, however, likely primes a genetic program that can be rapidly executed in response to JNK activation upon challenge. This notion is supported by the significant enrichment for functional GO terms associated with stress-related JNK signaling. Furthermore, Ets21c was shown to regulate genes that have been functionally linked to JNK signaling, including the autophagy-related gene 1 (atg1) and the insulin signaling intersecting-target (Jafrac1), or identified as JNK- and paraquat-responsive genes, such as the eukaryotic translation initiation factor 2α kinase (PEK), a thioredoxin-like protein (fax), or a glutathione S-transferase (gstD10). It is suggested that failure to accelerate intestinal regeneration and mount a robust cytoprotective response underlie the increased sensitivity of ets21c-deficient flies to oxidative stress. The capacity of Ets21c to confer cytoprotection but also trigger apoptosis is in line with the described roles of JNK signaling. It is proposed that the repertoire of Ets21c-regulated target genes and the biological outcome of Ets21c activation depends on the strength and duration of cellular stress and the signaling landscape in which Ets21c operates. Such a model would be in accordance with a showing that low stress levels can accelerate epithelial renewal in the absence of EC death due to moderate JNK induction that stimulates ISC division, while additional input from Hippo signaling accelerates apoptosis to prevent EC overcrowding (Mundorf, 2019).

In addition to its role in coordinating cellular behaviors within the intestine, Ets21c emerges as an important determinant of the adult intestinal healthspan and overall lifespan. Optimizing proliferative homeostasis in high-turnover tissues by, for example, moderate inhibition of insulin/IGF or JNK signaling activities has proven effective in prolonging the lifespan. As Ets21c represents a key effector of JNK in the adult gut and its knockdown reduced ISC proliferation, it is plausible that balanced intestinal function may contribute to the lifespan extension of unchallenged ets21c mutant flies. However, the use of a full-body ets21c mutant prevents drawing of a causal relation between the gut-specific role of Ets21c and longevity. The tissue- and cell-type-specific contribution of Ets21c to adult lifespan remains a question for future studies to address (Mundorf, 2019).

Finally, ets21c has been repeatedly picked up by gene expression profiling studies to be markedly increased in response to immune challenge, injury, oncogene activation, or aging. While functionally linked to JNK signaling in epithelial tumor models (Külshammer, 2015, Toggweiler, 2016), Ets21c has also been classified as an effector of EGFR/ERK signaling in the intestine based on the binding of Capicua, the EGFR/Ras/ERK-regulated transcriptional repressor, to the ets21c locus and upregulation of ets21c expression following Capicua loss (Jin, 2015). However, the functional evidence placing Ets21c downstream of EGFR/Ras/ERK signaling has been missing. The current data show that while knockdown of ets21c completely rescues the phenotypic consequences of JNK activation in the ISCs or ECs, it fails to mitigate the effects of activated EGFR/Ras/ERK signaling. Therefore, it is proposed that the regulation of ets21c levels results from an integration of positive and negative signaling inputs. This regulatory network includes Capicua, which acts as a gatekeeper of ets21c transcription. Such regulatory mechanisms ensure that JNK-Ets21c-mediated responses are fast but transient in supporting efficient tissue renewal while preventing chronic or excessive Ets21c activation, which drives tissue dysplasia and epithelial degeneration (Mundorf, 2019).

DamID profiling of dynamic Polycomb-binding sites in Drosophila imaginal disc development and tumorigenesis

Tracking dynamic protein-chromatin interactions in vivo is key to unravel transcriptional and epigenetic transitions in development and disease. However, limited availability and heterogeneous tissue composition of in vivo source material impose challenges on many experimental approaches. This study has adapted cell-type-specific DamID-seq profiling for use in Drosophila imaginal discs and make FLP/FRT-based induction accessible to GAL driver-mediated targeting of specific cell lineages. In a proof-of-principle approach, ubiquitous DamID expression was used to describe dynamic transitions of Polycomb-binding sites during wing imaginal disc development and in a scrib tumorigenesis model. Atf3 and Ets21C as novel Polycomb target genes involved in scrib tumorigenesis and suggest that target gene regulation by Atf3 and AP-1 transcription factors, as well as modulation of insulator function, plays crucial roles in dynamic Polycomb-binding at target sites. These findings by DamID-seq analysis of wing imaginal disc samples derived from 10 larvae. This study opens avenues for robust profiling of small cell population in imaginal discs in vivo and provides insights into epigenetic changes underlying transcriptional responses to tumorigenic transformation (La Fortezza, 2018).

The transcription factor Ets21C drives tumor growth by cooperating with AP-1

Tumorigenesis is driven by genetic alterations that perturb the signaling networks regulating proliferation or cell death. In order to block tumor growth, one has to precisely know how these signaling pathways function and interplay. This study has identified the transcription factor Ets21C as a pivotal regulator of tumor growth, and a new model is proposed of how Ets21C could affect this process. A depletion of Ets21C strongly suppressed tumor growth while ectopic expression of Ets21C further increased tumor size. It was confirmed that Ets21C expression is regulated by the JNK pathway, and Ets21C was shown to acts via a positive feed-forward mechanism to induce a specific set of target genes that is critical for tumor growth. These genes are known downstream targets of the JNK pathway, and their expression was demonstrated to not only depend on the transcription factor AP-1, but also on Ets21C, suggesting a cooperative transcriptional activation mechanism. Taken together this study shows that Ets21C is a crucial player in regulating the transcriptional program of the JNK pathway and enhances understanding of the mechanisms that govern neoplastic growth (Toggweiler, 2016).

This study has identified the transcription factor Ets21C as a crucial regulator of tumor growth and demonstrates that its expression is activated by the JNK pathway. Moreover, Ets21C was shown to regulate the expression of specific target genes that induce and sustain growth and invasiveness of RasV12 dlgRNAi tumors, possibly via a cooperation with AP-1 (Toggweiler, 2016).

The closest human orthologs of Ets21C ETS-related gene (ERG) and Friend leukemia virus-induced erythroleukemia 1 (FLI-1), have also been linked to tumorigenesis. ERG is overexpressed in acute myeloid leukemia (AML) and is associated with a poor prognosis, whereas higher FLI-1 expression has been detected in triple negative breast cancer or in metastatic melanoma. This study shows that Ets21C is not only sufficient to induce tumorigenesis, but required for tumor growth as a depletion of Ets21C strongly reduced tumor size. The smaller tumor size was accompanied by decreased levels of target genes known to drive tumor growth and malignancy, further underscoring the relevance of Ets21C. The finding that an Ets21C loss of function is blocking tumor growth so efficiently, suggests a more fundamental role for Ets21C in tumor growth than previously assumed. Kulshammer (2015) also tested an Ets21CRNAi in RasV12, scrib-/- tumors, but was not able to observe any remarkable effects besides a partial rescue of the pupariation delay that is commonly associated with neoplastic tumor growth in flies. An explanation for this discrepancy might simply be different strengths of the RNAi lines used (Toggweiler, 2016).

Loss of epithelial polarity caused by a depletion of dlg activates the JNK pathway that critically affects tumor growth. In agreement with previous studies, this study found elevated Ets21C transcript levels in RasV12 dlgRNAi tumors and showed that JNK signaling is required for the upregulation. The JNK pathway activates the AP-1 transcription factor, which in its prototypical form is a dimer consisting of Jun and Fos proteins. In the RasV12, scrib-/- context, Fos has been described as the main effector of the JNK pathway, as a depletion diminishes tumor growth and abrogates induction of target genes, while no such effect have been observed for Jun. However, the current data point towards at least a partial role for Jun activating target genes, since a knockdown of Jun reduced expression of Ets21C and Mmp1 (Toggweiler, 2016).

Given the strong effects that Ets21C exerts on tumor growth, attempts were made to identify Ets21C dependent target genes that could explain this phenomenon. Transcriptional profiles revealed an upregulation of Mmp1, Pvf1 and upd1 in tumors that overexpressed Ets21C and downregulated in tumors depleted of Ets21C. Mmp1, Pvf1 and upd1 have previously been shown to be induced by the JNK pathway and to be essential for tumor growth and invasion. Besides upd1, also upd2 and upd3 have been reported to be induced in neoplastic tumors in a JNK pathway dependent manner. In agreement with these reports, an induction was observed of upd2 and upd3 in RasV12 dlgRNAi Ets21C tumors, but to a lesser extent than upd1. This might on the one hand depend on the use of dlg instead of scrib a stronger activation of Upd cytokines has been found in scrib mutant wing discs compared to dlg mutants. On the other hand, additional factors might influence transcriptional activation such as the presence of RasV12, the exact timing of sample collection or the genetic system used. Furthermore, the strength of induction might not necessarily correlate with the effect on tumor growth. Although upd3 exhibits the strongest upregulation in neoplastic tumors, it has been shown that co-expression of Upd1 and Upd2 together with RasV12 leads to much larger tumors than the combination of RasV12 and Upd3 (Toggweiler, 2016).

While Ets21C is able to stimulate the expression of the JNK downstream targets Mmp1, Pvf1 and upd1, no obvious regulation was observed of canonical JNK pathway targets such as the phosphatase puc or the apoptosis inducers hid or rpr, suggesting that Ets21C only regulates a specific set of JNK pathway targets. In contrast, a previous study has described a slight upregulation of puc in RasV12 Ets21C tumors. Trivial differences such as when the samples were collected, transgene strength, or genetic background could account for this difference. For example, an elevation of puc expression could also originate from an indirect increase in JNK activity due to stresses in an older, larger tumor. These results are entirely consistent with the model that Ets21C can activate certain JNK targets in a context dependent manner (Toggweiler, 2016).

Finally, it was asked if and how Ets21C regulates transcriptional outputs of the JNK pathway. (1) Ets21C could interact with Jun and/or Fos. (2) Ets21C could activate or interact with an unknown factor that feeds back on the JNK pathway. (3) Ets21C could use a combination of both (1) and (2). This study showed genetically that in RasV12 dlgRNAi tumors the effects of Ets21C, both phenotypically and on a target genes level, fully depend on an active JNK pathway. If the latter is blocked, for example, by co-expressing BskDN with RasV12 dlgRNAi and Ets21CHA, tumors remain small and there is no induction of Mmp1, Pvf1 or upd1. These results support all possibilities (1)-(3), but exclude an autonomous function of Ets21C. A physical interaction between Ets21C and Jun and Fos has previously been proposed based on large scale mass-spectroscopy data. This study shows that HA-tagged Ets21C does indeed bind FLAG-tagged Jun or Fos, showing that the proteins could interact physically to regulate target gene expression. The binding is consistent with studies in mammalian systems that have shown a physical interaction between AP-1 and ETS proteins including ERG, the mammalian homolog of Ets21C. In agreement with a cooperative transcriptional activation, Ets21C and AP-1 binding sites were found co-occurring in the putative regulatory regions of the analyzed target genes. It is therefore thought that Ets21C likely activates target genes via a cooperation with AP-1. However, additional factors that are activated by the JNK pathway might contribute to target gene activation as well and could explain why Ets21C activates certain JNK downstream target genes and others not (Toggweiler, 2016).

In summary, this study shows that Ets21C fulfills a crucial role in the regulation of neoplastic tumor growth as a loss of function critically reduced tumor growth, whereas an excess of Ets21C further increased tumor size. While Ets21C was previously accredited only with a role in fine-tuning the transcriptional program of neoplastic tumors, the current results point towards a more fundamental role as activator of a specific set of target genes that drive tumor growth and invasion (Toggweiler, 2016).

EGFR/Ras signaling controls Drosophila intestinal stem cell proliferation via Capicua-regulated genes

Epithelial renewal in the Drosophila intestine is orchestrated by Intestinal Stem Cells (ISCs). Following damage or stress the intestinal epithelium produces ligands that activate the epidermal growth factor receptor (EGFR) in ISCs. This promotes their growth and division and, thereby, epithelial regeneration. This study demonstrates that the HMG-box transcriptional repressor, Capicua (Cic), mediates these functions of EGFR signaling. Depleting Cic in ISCs activated them for division, whereas overexpressed Cic inhibited ISC proliferation and midgut regeneration. Epistasis tests showed that Cic acted as an essential downstream effector of EGFR/Ras signaling, and immunofluorescence showed that Cic's nuclear localization was regulated by EGFR signaling. ISC-specific mRNA expression profiling and DNA binding mapping using DamID indicated that Cic represses cell proliferation via direct targets including string (Cdc25), Cyclin E, and the ETS domain transcription factors Ets21C and Pointed (pnt). pnt was required for ISC over-proliferation following Cic depletion, and ectopic pnt restored ISC proliferation even in the presence of overexpressed dominant-active Cic. These studies identify Cic, Pnt, and Ets21C as critical downstream effectors of EGFR signaling in Drosophila ISCs (Jin 2015).

It is well established that EGFR signaling is essential to drive ISC growth and division in the fly midgut. However, the precise mechanism via which this signal transduction pathway activates ISCs has remained a matter of inference from experiments with other cell types. Moreover, despite a vast literature on the pathway and ubiquitous coverage in molecular biology textbooks, the mechanisms of action of the pathway downstream of the MAPK are not well understood for any cell type. From this study, a model is proposed (see Model for Cic control of Drosophila ISC proliferation). Multiple EGFR ligands and Rhomboid proteases are induced in the midgut upon epithelial damage, which results in the activation of the EGFR, RAS, RAF, MEK, and MAPK in ISCs. MAPK phosphorylates Cic in the nucleus, which causes it to dissociate from regulatory sites on its target genes and also translocate to the cytoplasm. This allows the de-repression of target genes, which may then be activated for transcription by factors already present in the ISCs. The critical Cic target genes identified in this study include the cell cycle regulators stg (Cdc25) and Cyclin E, which in combination are sufficient to drive dormant ISCs through S and M phases, and pnt and Ets21C, ETS-type transcriptional activators that are required and sufficient for ISC activation (Jin 2015).

Upon damage, activated EGFR signaling mediates activation of ERK, which phosphorylates Cic, and relocates it to the cytoplasm. As a result, stg, CycE, Ets21C and pnt transcription are relieved from Cic repression, and induce ISC proliferation (Jin 2015).

Although this study found more than 2000 Cic binding sites in the ISC genome, not all of the genes associated with these sites were significantly upregulated, as assayed by RNA-Seq, upon Cic depletion. One possible explanation for this is that Cic binding sites from DamID-Seq were also associated with other types of transcription units (miRNAs, snRNAs, tRNAs, rRNAs, lncRNAs) that were not scored for activation by the RNA-Seq analysis. Indeed a survey of the Cic binding site distributions suggests this. This might classify some binding sites as non-mRNA-associated. However, it is also possible that many Cic target genes may require activating transcription factors that are not expressed in ISCs. Such genes might not be strongly de-repressed in the gut upon Cic depletion (Jin 2015).

In other Drosophila cells MAPK phosphorylation is thought to directly inactivate the ETS domain repressor Yan, and to directly activate the ETS domain transcriptional activator Pointed P2 (PNTP2). In fact Pnt and Yan have been shown to compete for common DNA binding sites on their target genes. Thus, previous studies proposed a model of transcriptional control by MAPK based solely on post-translational control of the activity of these ETS factors. However, this study found that Pnt and Ets21C were transcriptionally induced by MAPK signaling, and could activate ISCs if overexpressed, and that depleting yan or pntP2 had no detectable proliferation phenotype. In addition, overexpression of PNTP2 was sufficient to trigger ISC proliferation, suggesting either that basal MAPK activity is sufficient for its post-translational activation, or that PNTP2 phosphorylation is not obligatory for activity. Moreover, pntP2 loss of function mutant ISC clones had no deficiency in growth even after inducing proliferation by P.e. infection, which increases MAPK signaling. These observations indicate that the direct MAPKā†’PNTP2 phospho-activation pathway is not uniquely or specifically required for ISC proliferation. These results suggest instead that transcriptional activation of pnt and Ets21c via MAPK-dependent loss of Cic-mediated repression is the predominant mode of downstream regulation by MAPK in midgut ISCs (Jin 2015).

In addition to activating ISCs for division, EGFR signaling activates them for growth. Previous studies showed loss of EGFR signaling prevented ISC growth and division, and that ectopic RasV12 expression could accelerate the growth not only of ISCs but also post-mitotic enteroblasts. Similarly, this study shows that loss of cic caused ISC clones to grow faster than controls, by increasing cell number as well as cell size. For instance, increased size of GFP+ ISCs and EBs was observed when cic-RNAi was induced by the esgts or esgtsF/O systems. Therefore, in a search for Cic target genes probable growth regulatory genes such as Myc, Cyclin D, the Insulin/TOR components InR, PI3K, S6K and Rheb, Hpo pathway components, and the loci encoding rRNA, tRNAs and snRNAs were specifically checked. It was found that Cic bound to the InR, Akt1, Rheb, Src42A and Yki loci. However, of these only InR mRNA was significantly upregulated in Cic-depleted progenitor cells. In surveying the non-protein coding genome, it was found that Cic had binding sites in many loci encoding tRNA, snRNA, snoRNA and other non-coding RNAs, though not in the 28S rRNA or 5S rRNA genes. Due to the method used for RNA-Seq library production, RNA expression profiling experiments could not detect expression of these loci, and so it remains to be tested whether Cic may regulate some of those non-coding RNA's transcription to control cell growth. It is also possible that Cic controls cell growth regulatory target genes indirectly, for instance via its targets Ets21C and Pnt, which are also strong growth promoters in the midgut. But given that no conclusive model can be drawn from the data regarding how Cic restrains growth, one must consider the possibility that ERK signaling stimulates cell growth via non-transcriptional mechanisms, as proposed by several studies (Jin 2015).

The critical role of Cic as a negative regulator of cell proliferation in the fly midgut is consistent with its tumor suppressor function in mammalian cancer development. Also consistent with the current findings are the observations that the ETS transcription factors ETV1 and ETV5 are upregulated in sarcomas that express CIC-DUX, an oncogenic fusion protein that functions as a transcriptional activator, and that knockdown of CIC induces the transcription of ETV1, ETV4 and ETV5 in melanoma cells. Moreover the transcriptional regulation by ETS transcription factors is important in human cancer development. Their expression is induced in many tumors and cancer cell lines. For example, ERG, ETV1, and ETV4 can be upregulated in prostrate cancers, and ETV1 is upregulated in post gastrointestinal stromal tumors and in more than 40% of melanomas. In addition, the mRNA expression of these ETS genes was correlated with ERK activity in melanoma and colon cancer cell lines with activating mutations in BRAF (V600E), such that their expression decreased upon MEK inhibitor treatment. Furthermore, overexpression of the oncogenic ETS proteins ERG or ETV1 in normal prostate cells can activate a Ras/MAPK-dependent gene expression program in the absence of ERK activation. These cancer studies imply that there is an unknown factor that links Ras/Mapk activity to the expression of ETS factors, and that some of the human ETS factors might act without MAPK phosphorylation, as does Drosophila PntP1. Combining the knowledge of Cic with what was previously known about CIC in tumor development, it is proposed that CIC is the unknown factor that regulates ETS transcription factors in Ras/MAKP-activated human tumors (Jin 2015).

In summary, this study has elucidated a mechanism wherein Cic controls the expression of the cell cycle regulators stg (Cdc25) and Cyclin E, along with the Ets transcription factor Pnt, and perhaps also Ets21C, by directly binding to regulatory sites in their promoters and introns. Using genetic tests it was shown that these interactions are meaningful for regulating stem cell proliferation. Therefore, it is suggested that human CIC may also lead to the transcriptional induction of cell cycle genes and ETS transcription factors in RAS/MAPK activated- or loss-of-function-CIC tumors such as brain or colorectal cancers (Jin 2015).

Interplay among Drosophila transcription factors Ets21c, Fos and Ftz-F1 drives JNK-mediated tumor malignancy

This study defines TF network that triggers an abnormal gene expression program promoting malignancy of clonal tumors, generated in Drosophila imaginal disc epithelium by gain of oncogenic Ras (RasV12) and loss of the tumor suppressor Scribble (scrib1). Malignant transformation of the rasV12scrib1 tumors requires TFs of distinct families, namely the bZIP protein Fos, the ETS-domain factor Ets21c and the nuclear receptor Ftz-F1, all acting downstream of Jun-N-terminal kinase (JNK). Depleting any of the three TFs improves viability of tumor-bearing larvae, and this positive effect can be enhanced further by their combined removal. Although both Fos and Ftz-F1 synergistically contribute to rasV12scrib1 tumor invasiveness, only Fos is required for JNK-induced differentiation defects and Matrix metalloprotease (MMP1) upregulation. In contrast, the Fos-dimerizing partner Jun is dispensable for JNK to exert its effects in rasV12scrib1 tumors. Interestingly, Ets21c and Ftz-F1 are transcriptionally induced in these tumors in a JNK- and Fos-dependent manner, thereby demonstrating a hierarchy within the tripartite TF network, with Fos acting as the most upstream JNK effector. Of the three TFs, only Ets21c can efficiently substitute for loss of polarity and cooperate with Ras(V12) in inducing malignant clones that, like rasV12scrib1 tumors, invade other tissues and overexpress MMP1 and the Drosophila insulin-like peptide 8 (Dilp8). While rasV12ets21c tumors require JNK for invasiveness, the JNK activity is dispensable for their growth. In conclusion, this study delineates both unique and overlapping functions of distinct TFs that cooperatively promote aberrant expression of target genes, leading to malignant tumor phenotypes. (Kulshammer, 2015).

Genome-wide transcriptome profiling in the Drosophila epithelial tumor model has generated a comprehensive view of gene expression changes induced by defined oncogenic lesions that cause tumors of an increasing degree of malignancy. These data allowed discovery of how a network of collaborating transcription factors confers malignancy to RasV12scrib1 tumors (Kulshammer, 2015).

This study revealed that the response of transformed RasV12scrib1 epithelial cells is more complex in comparison to those with activated RasV12 alone with respect to both the scope and the magnitude of expression of deregulated genes. Aberrant expression of more than half of the genes in RasV12scrib1 tumors requires JNK activity, highlighting the significance of JNK signaling in malignancy. Importantly, the tumor-associated, JNK-dependent transcripts cluster with biological functions and processes that tightly match the phenotypes of previously described tumor stages. Furthermore, the RasV12scrib1 transcriptome showed significant overlap (27% upregulated and 15% downregulated genes) with microarray data derived from mosaic EAD in which tumors were induced by overexpressing the BTB-zinc finger TF Abrupt (Ab) in scrib1 mutant clones as well as with a transcriptome of scrib1 mutant wing discs. It is proposed that 429 misregulated transcripts (e.g. cher, dilp8, ets21c, ftz-f1, mmp1, upd), shared among all the three data sets irrespective of epithelial type (EAD versus wing disc) or cooperating lesion (RasV12 or Ab), represent a 'polarity response transcriptional signature' that characterizes the response of epithelia to tumorigenic polarity loss. Genome-wide profiling and comparative transcriptome analyses thus provide a foundation to identify novel candidates that drive and/or contribute to tumor development and malignancy while unraveling their connection to loss of polarity and JNK signaling (Kulshammer, 2015).

In agreement with a notion of combinatorial control of gene expression by an interplay among multiple TFs, this study identified overrepresentation of cis-acting DNA elements for STAT, GATA, bHLH, ETS, BTB, bZIP factors and NRs in genes deregulated in RasV12scrib1 mosaic EAD, implying that transcriptome anomalies result from a cross-talk among TFs of different families. Many of the aberrantly expressed genes contained binding motifs for AP-1, Ets21c and Ftz-F1, indicating that these three TFs may regulate a common set of targets and thus cooperatively promote tumorigenesis. This is consistent with the occurrence of composite AP-1-NRRE (nuclear receptor response elements), ETS-NRRE and ETS-AP-1 DNA elements in the regulatory regions of numerous human cancer-related genes, such as genes for cytokines, MMPs (e.g., stromelysin, collagenase) and MMP inhibitors (e.g., TIMP) (Kulshammer, 2015).

Interestingly, Drosophila ets21c and ftz-f1 gene loci themselves contain AP-1 motifs and qualify as polarity response transcriptional signature transcripts. Indeed, this study has detected JNK- and Fos-dependent upregulation of ets21c and ftz-f1 mRNAs in RasV12scrib1 tumors. While JNK-mediated control of ftz-f1 transcription has not been reported previously, upregulation of ets21c in the current tumor model is consistent with JNK requirement for infection-induced expression of ets21c mRNA in Drosophila S2 cells and in vivo. Based on these data, it is proposed that Ftz-F1 and Ets21c are JNK-Fos-inducible TFs that together with AP-1 underlie combinatorial transcriptional regulation and orchestrate responses to cooperating oncogenes. Such an interplay between AP-1 and Ets21c is further supported by a recent discovery of physical interactions between Drosophila Ets21c and the AP-1 components Jun and Fos. Whether regulatory interactions among AP-1, Ets21c and Ftz-F1 require their direct physical contact and/or the presence of composite DNA binding motifs of a particular arrangement to control the tumor-specific transcriptional program remains to be determined (Kulshammer, 2015).

Importantly, some of the corresponding DNA elements, namely AP-1 and STAT binding sites, have recently been found to be enriched in regions of chromatin that become increasingly accessible in RasV12scrib1 mosaic EAD relative to control. This demonstrates that comparative transcriptomics and open chromatin profiling using ATAC-seq and FAIRE-seq are suitable complementary approaches for mining the key regulatory TFs responsible for controlling complex in vivo processes, such as tumorigenesis (Kulshammer, 2015).

The prototypical form of AP-1 is a dimer comprising Jun and Fos proteins. In mammals, the Jun proteins occur as homo- or heterodimers, whereas the Fos proteins must interact with Jun in order to bind the AP-1 sites. In contrast to its mammalian orthologs, the Drosophila Fos protein has been shown to form a homodimer capable of binding to and activating transcription from an AP-1 element, at least in vitro (Kulshammer, 2015).

The role of individual AP-1 proteins in neoplastic transformation and their involvement in pathogenesis of human tumors remain somewhat elusive. While c-Jun, c-Fos and FosB efficiently transform mammalian cells in vitro, only c-Fos overexpression causes osteosarcoma formation, whereas c-Jun is required for development of chemically induced skin and liver tumors in mice. In contrast, JunB acts as a context-dependent tumor suppressor. Thus, cellular and genetic context as well as AP-1 dimer composition play essential roles in dictating the final outcome of AP-1 activity in tumors (Kulshammer, 2015).

This study shows that, similar to blocking JNK with its dominant-negative form, Bsk, removal of Fos inhibits ets21c and ftz-f1 upregulation, suppresses invasiveness, improves epithelial organization and differentiation within RasV12scrib1 tumors and allows larvae to pupate. Strikingly, depletion of Jun had no such tumor-suppressing effects. It is therefore concluded that in the malignant RasV12scrib1 tumors, Fos acts independently of Jun, either as a homodimer or in complex with another, yet unknown partner. A Jun-independent role for Fos is further supported by additional genetic evidence. Fos, but not Jun, is involved in patterning of the Drosophila endoderm and is required for expression of specific targets, e.g., misshapen (msn) and dopa decarboxylase (ddc), during wound healing. Future studies should establish whether the JNK-responsive genes containing AP-1 motifs, identified in this study, are indeed regulated by Fos without its 'canonical' partner (Kulshammer, 2015).

The current data identify Fos as a key mediator of JNK-induced MMP1 expression and differentiation defects in RasV12scrib1 tumors. Only Fos inhibition caused clear suppression of MMP1 levels and restoration of neurogenesis within clonal EAD tissue, thus mimicking effects of JNK inhibition. Improved differentiation and reduced invasiveness are, however, not sufficient for survival of animals to adulthood, because interfering with Fos function in RasV12scrib1 clones always resulted in pupal lethality (Kulshammer, 2015).

The systems approach of this paper, followed by genetic experiments, identified Ets21c and Ftz-F1 as being essential for RasV12scrib1-driven tumorigenesis. It was further shown that mutual cooperation of both of these TFs with Fos is required to unleash the full malignancy of RasV12scrib1 tumors (Kulshammer, 2015).

TFs of the ETS-domain family are key regulators of development and homeostasis in all metazoans, whereas their aberrant activity has been linked with cancer. ets21c encodes the single ortholog of human Friend leukemia insertion1 (FLI1) and ETS-related gene (ERG) that are commonly overexpressed or translocated in various tumor types. While FLI1 is considered pivotal to development of Ewing's sarcoma, ERG has been linked to leukemia and prostate cancer. As for Ftz-F1 orthologs, the human liver receptor homolog-1 (LRH-1) has been associated with colonic, gastric, breast and pancreatic cancer, whereas steroidogenic factor 1 (SF-1) has been implicated in prostate and testicular cancers and in adrenocortical carcinoma. However, the molecular mechanisms underlying oncogenic activities of either the ERG/FLI1 or the SF-1/LRH-1 proteins are not well understood (Kulshammer, 2015).

This study shows that removal of Ftz-F1 markedly suppressed invasiveness of RasV12scrib1 tumors, restoring the ability of tumor-bearing larvae to pupate. Additionally, and in contrast to Fos, Ftz-F1 inhibition also partly reduced tumor growth in the third-instar EAD and allowed emergence of adults with enlarged, rough eyes composed predominantly of non-clonal tissue. The reduced clonal growth coincided with downregulation of the well-established Yki target, expanded, implicating Ftz-F1 as a potential novel growth regulator acting on the Hpo/Yki pathway. It is further speculated that reduced viability of RasV12scrib1ftz-f1RNAi clones and induction of non-autonomous compensatory proliferation by apoptotic cells during the pupal stage could explain the enlargement of the adult eyes. The precise mechanism underlying compromised growth and invasiveness of RasV12scrib1ftz-f1RNAi tumors and improved survival of the host remains to be identified (Kulshammer, 2015).

In contrast, effects of Ets21cLONG knockdown in RasV12scrib1 tumors appeared moderate relative to the clear improvement conferred by either Fos or Ftz-F1 elimination. ets21cLONG RNAi neither reduced tumor mass nor suppressed invasiveness, and pupation was rescued only partly. However, unlike ftz-f1RNAi, ets21cLONG RNAi significantly reduced expression of dilp8 mRNA. Based on abundance of Ets21c binding motifs in the regulatory regions of tumor-associated genes and the normalized expression of >20% of those genes upon removal of Ets21c, it is further suggested that Ets21c acts in RasV12scrib1 tumors to fine-tune the tumor gene-expression signature (Kulshammer, 2015).

Dilp8 is known to be secreted by damaged, wounded or tumor-like tissues to delay the larval-to-pupal transition. This study has corroborated the role of JNK in stimulating dilp8 expression in RasV12scrib1 tumor tissue, and has further implicated Ets21c and Fos as novel regulators of dilp8 downstream of JNK. However, the data also show that elevated dilp8 transcription per se is not sufficient to delay metamorphosis. Unlike the permanent larvae bearing RasV12scrib1 tumors, those with RasV12scrib1ftz-f1RNAi tumors pupated despite the excessive dilp8 mRNA. Likewise, pupation was not blocked by high dilp8 levels in larvae bearing EAD clones overexpressing Abrupt. As Dilp8 secretion appears critical for its function, it is proposed that loss of Ftz-F1 might interfere with Dilp8 translation, post-translational processing or secretion (Kulshammer, 2015).

Consistent with the individual TFs having unique as well as overlapping functions in specifying properties of RasV12scrib1 tumors, knocking down pairwise combinations of the TFs had synergistic effects on tumor suppression compared with removal of single TF. This evidence supports the view that malignancy is driven by a network of cooperating TFs, and elimination of several tumor hallmarks dictated by this network is key to animal survival. An interplay between AP-1, ETS-domain TFs and NRs is vital for development. For example, the ETS-factor Pointed has been shown to cooperate with Jun to promote R7 photoreceptor formation in the Drosophila adult eye. In mosquitoes, synergistic activity of another ETS-factor, E74B, with the ecdysone receptor (EcR/USP) promotes vitellogenesis. It is thus proposed that tumors become malignant by hijacking the developmental mechanism of combinatorial control of gene activity by distinct TFs (Kulshammer, 2015).

Despite the minor impact of ets21cLONG knockdown on suppressing RasV12scrib1 tumors, Ets21cLONG is the only one of the tested TFs that was capable of substituting for loss of scrib in inducing malignant clonal overgrowth when overexpressed with oncogenic RasV12 in EAD. While invasiveness of such RasV12ets21cLONG tumors required JNK activity, JNK signaling appeared dispensable for tumor growth. Importantly, the overgrowth of RasV12ets21cLONG tumors was primarily independent of a prolonged larval stage, because dramatic tumor mass expansion was detected already on day 6 AEL. How cooperativity between Ets21cLONG and RasV12 ensures sufficient JNK activity and the nature of the downstream effectors driving tumor overgrowth remain to be determined. In contrast, co-expression of either Ftz-F1 or Fos with RasV12 resulted in a non-invasive, RasV12-like hyperplastic phenotype (Kulshammer, 2015).

Why does Ets21cLONG exert its oncogenic potential while Fos and Ftz-F1 do not? Simple overexpression of a TF may not be sufficient, because many TFs require activation by a post-translational modification (e.g., phosphorylation), interaction with a partner protein and/or binding of a specific ligand. Full activation of Fos in response to a range of stimuli is achieved through hyperphosphorylation by mitogen-activated protein kinases (MAPKs), including ERK and JNK. Indeed, overexpression of a FosN-Ala mutated form that cannot be phosphorylated by JNK was sufficient to phenocopy fos deficiency, indicating that Fos must be phosphorylated by JNK in order to exert its oncogenic function. Consistent with the current data, overexpression of FosN-Ala partly restored polarity of lgl mutant EAD cells. It is therefore conclude that the tumorigenic effect of Fos requires a certain level of JNK activation, which is lacking in EAD co-expressing Fos with RasV12. Nevertheless, the absence of an unknown Fos-interacting partner cannot be excluded (Kulshammer, 2015).

Interestingly, MAPK-mediated phosphorylation also greatly enhances the ability of SF-1 and ETS proteins to activate transcription. Two potential MAPK sites can be identified in the hinge region of Ftz-F1, although their functional significance is unknown. Whether Ets21c or Ftz-F1 requires phosphorylation and how this would impact their activity in the tumor context remains to be determined. Genetic experiments demonstrate that at least the overgrowth of RasV12ets21cLONG tumors does not require Ets21c phosphorylation by JNK (Kulshammer, 2015).

In addition, previous crystallography studies revealed the presence of phosphoinositides in the ligand binding pocket of LHR-1 and SF-1 and showed their requirement for the NR transcriptional activity. Although developmental functions of Drosophila Ftz-F1 seem to be ligand independent, it is still possible that Ftz-F1 activity in the tumor context is regulated by a specific ligand. An effect of Ftz-F1 SUMOylation cannot be ruled out (Kulshammer, 2015).

In summary, this work demonstrates that malignant transformation mediated by RasV12 and scrib loss depends on MAPK signaling and at least three TFs of different families, Fos, Ftz-F1 and Ets21c. While their coordinated action ensures precise transcriptional control during development, their aberrant transcriptional (Ets21c, Ftz-F1) and/or post-translational (Fos, Ftz-F1, Ets21c) regulation downstream of the cooperating oncogenes contributes to a full transformation state. The data implicate Fos as a primary nuclear effector of ectopic JNK activity downstream of disturbed polarity that controls ets21c and ftz-f1 expression. Through combinatorial interactions on overlapping sets of target genes and acting on unique promoters, Fos, Ftz-F1 and Ets21c dictate aberrant behavior of RasV12scrib1 tumors. Although originally described in Drosophila, detrimental effects of cooperation between loss of Scrib and oncogenic Ras has recently been demonstrated in mammalian tumor models of prostate and lung cancer. This study and further functional characterization of complex TF interactions in the accessible Drosophila model are therefore apt to provide important insight into processes that govern cancer development and progression in mammals (Kulshammer, 2015).

How the fly balances its ability to combat different pathogens

Health is a multidimensional landscape. If the host alone is considered, there are many outputs that are of interest: evolutionary fitness determining parameters like fecundity, survival and pathogen clearance as well as medically important health parameters like sleep, energy stores and appetite. Hosts use a variety of effector pathways to fight infections and these effectors are brought to bear differentially. Each pathogen causes a different disease as they have distinct virulence factors and niches; they each warp the health landscape in unique ways. Therefore, mutations affecting immunity can have complex phenotypes and distinct effects on each pathogen. This study describes how two components of the fly's immune response, melanization and phagocytosis, contribute to the health landscape generated by the transcription factor ets21c and its putative effector, the signaling molecule wntD (CG8458). To probe the landscape, flies were infected with two pathogens: Listeria monocytogenes, which primarily lives intracellularly, and Streptococcus pneumoniae, which is an extracellular pathogen. Using the diversity of phenotypes generated by these mutants, it is proposed that survival during a L. monocytogenes infection is mediated by a combination of two host mechanisms: phagocytic activity and melanization; while survival during a S. pneumoniae infection is determined by phagocytic activity. In addition, increased phagocytic activity is beneficial during S. pneumoniae infection but detrimental during L. monocytogenes infection, demonstrating an inherent trade-off in the immune response (Chambers, 2012).


REFERENCES

Search PubMed for articles about Drosophila Ets21C

Biteau, B., Hochmuth, C. E. and Jasper, H. (2008). JNK activity in somatic stem cells causes loss of tissue homeostasis in the aging Drosophila gut. Cell Stem Cell 3(4): 442-455. PubMed ID: 18940735

Biteau, B., Karpac, J., Supoyo, S., Degennaro, M., Lehmann, R. and Jasper, H. (2010). Lifespan extension by preserving proliferative homeostasis in Drosophila. PLoS Genet 6(10): e1001159. PubMed ID: 20976250

Blanco, E., Ruiz-Romero, M., Beltran, S., Bosch, M., Punset, A., Serras, F. and Corominas, M. (2010). Gene expression following induction of regeneration in Drosophila wing imaginal discs. Expression profile of regenerating wing discs. BMC Dev Biol 10: 94. PubMed ID: 20813047

Boutros, M., Agaisse, H. and Perrimon, N. (2002). Sequential activation of signaling pathways during innate immune responses in Drosophila. Dev Cell 3(5): 711-722. PubMed ID: 12431377

Broderick, N. A., Buchon, N. and Lemaitre, B. (2014). Microbiota-induced changes in drosophila melanogaster host gene expression and gut morphology. MBio 5(3): e01117-01114. PubMed ID: 24865556

Chambers, M. C., Lightfield, K. L. and Schneider, D. S. (2012). How the fly balances its ability to combat different pathogens. PLoS Pathog 8(12): e1002970. PubMed ID: 23271964

Guo, L., Karpac, J., Tran, S. L. and Jasper, H. (2014). PGRP-SC2 promotes gut immune homeostasis to limit commensal dysbiosis and extend lifespan. Cell 156(1-2): 109-122. PubMed ID: 24439372

Jin, Y., Ha, N., Fores, M., Xiang, J., Glasser, C., Maldera, J., Jimenez, G. and Edgar, B. A. (2015). EGFR/Ras signaling controls Drosophila intestinal stem cell proliferation via Capicua-regulated genes. PLoS Genet 11: e1005634. PubMed ID: 26683696

Kulshammer, E., Mundorf, J., Kilinc, M., Frommolt, P., Wagle, P. and Uhlirova, M. (2015). Interplay among Drosophila transcription factors Ets21c, Fos and Ftz-F1 drives JNK-mediated tumor malignancy. Dis Model Mech 8(10): 1279-1293. PubMed ID: 26398940

La Fortezza, M., Grigolon, G., Cosolo, A., Pindyurin, A., Breimann, L., Blum, H., van Steensel, B. and Classen, A. K. (2018). DamID profiling of dynamic Polycomb-binding sites in Drosophila imaginal disc development and tumorigenesis. Epigenetics Chromatin 11(1): 27. PubMed ID: 29871666

Mundorf, J., Donohoe, C. D., McClure, C. D., Southall, T. D. and Uhlirova, M. (2019). Ets21c governs tissue renewal, stress tolerance, and aging in the Drosophila intestine. Cell Rep 27(10): 3019-3033.e3015. PubMed ID: 31167145

Patterson, R. A., Juarez, M. T., Hermann, A., Sasik, R., Hardiman, G. and McGinnis, W. (2013). Serine proteolytic pathway activation reveals an expanded ensemble of wound response genes in Drosophila. PLoS One 8(4): e61773. PubMed ID: 23637905

Toggweiler, J., Willecke, M. and Basler, K. (2016). The transcription factor Ets21C drives tumor growth by cooperating with AP-1. Sci Rep 6: 34725. PubMed ID: 27713480


Biological Overview

date revised: 12 July 2019

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