|
Gene name - archipelago
Synonyms - Cytological map position - 64A11--12 Function - signaling Keywords - component of SCF ubiquitin ligase complex, eye, oogenesis, growth |
Symbol - ago
FlyBase ID: FBgn0041171 Genetic map position - 3L Classification - F-box and WD repeat protein Cellular location - cytoplasmic |
The Myc oncoprotein is an important regulator of cellular growth in metazoan organisms. Its levels and activity are tightly controlled in vivo by a variety of mechanisms. In normal cells, Myc protein is rapidly degraded, but the mechanism of its degradation is not well understood. Genetic and biochemical evidence is presented that Archipelago (Ago), the F box component of an SCF-ubiquitin ligase and the Drosophila ortholog of a human tumor suppressor, negatively regulates the levels and activity of Drosophila Myc (dMyc) protein in vivo. Mutations in archipelago (ago) result in strongly elevated dMyc protein levels and increased tissue growth. Genetic interactions indicate that ago antagonizes dMyc function during development. Archipelago binds dMyc and regulates its stability, and the ability of Ago to bind dMyc in vitro correlates with its ability to inhibit dMyc accumulation in vivo. These data indicate that archipelago is an important inhibitor of dMyc in developing tissues. Because archipelago can also regulate Cyclin E levels and Notch activity, these results indicate how a single F box protein can be responsible for the degradation of key components of multiple pathways that control growth and cell cycle progression (Moberg, 2004).
myc genes encode basic-helix-loop-helix-zipper (bHLHZ) domain transcription factors that dimerize with Max family proteins to promote cell growth and proliferation in metazoan organisms. The Myc-Max complex is implicated in the transcriptional regulation of many genes that are required for cell growth and metabolism; such genes include those for translation initiation factors and ribosomal components. The role of Myc in promoting growth is likely to contribute to its role as an oncoprotein in a wide variety of human tumor types. myc overexpression also promotes tumorigenesis in mice and zebrafish, indicating that the oncogenic properties of myc genes are conserved in other organisms (Moberg, 2004 and references therein).
Deregulation of mammalian Myc in cancer occurs by a variety of mechanisms. In some cancers, notably lymphomas, mutations found within the Myc protein have been shown to increase its stability (Salghetti, 1999; Gregory, 2000). Myc protein is normally turned over rapidly in vivo and in cultured cells has a half-life of 20-30 min, and several studies have shown that Myc protein is subject to ubiquitin-dependent proteasomal degradation (Grandori, 2000). Ubiquitination of Myc in turn appears to be regulated by phosphorylation at two distinct sites in the protein's amino-terminal portion, Threonine 58 (Thr58) and Serine 62 (Ser62). Evidence suggests that MAP kinase mediates phosphorylation of Ser62 and that this stabilizes c-Myc. Phospho-Ser62 may be required for subsequent phosphorylation of Thr58 by glycogen-synthase kinase 3 (GSK3: Drosophila homolog, Shaggy), which promotes the ubiquitination and degradation of c-Myc (Gregory, 2003). Significantly, Thr58Ile is the most common c-Myc mutation in Burkitt's lymphoma and is known to stabilize Myc considerably (Salghetti, 1999; Gregory, 2000). These observations suggest that phosphorylation of Thr58 by GSK3 generates a motif that facilitates the interaction of Myc with a ubiquitin ligase that restricts Myc levels and activity in vivo. Currently, the identity of the ubiquitin ligase that promotes Myc degradation has not been firmly established in any organism (Moberg, 2004 and references therein).
The Drosophila F box protein Archipelago (Ago) has been implicated in the degradation of Drosophila Myc (dMyc). Ago binds dMyc, and impairment of Ago function in vivo stabilizes dMyc, resulting in markedly elevated Myc levels, and promotes cell growth. Recent evidence indicates that the Fbw7/hCDC4 tumor suppressor protein, which is the human ortholog of Ago, also inhibits c-Myc accumulation by promoting its degradation (Welcker, 2004). Because Ago proteins also regulate Cyclin E levels (Moberg, 2001; Strohmaier, 2001; Koepp, 2001), and Notch pathway activity (Moberg, unpublished observations and reviewed in Justice, 2002), these findings suggest a mechanism by which the levels of Cyclin E and dMyc and the activity of the Notch pathway can be coordinately regulated by a shared degradation pathway (Moberg, 2004).
Thus Ago, which functions as the substrate-specificity subunit of an SCFAgo ubiquitin ligase, regulates the levels of the growth-promoting transcription factor dMyc in developing Drosophila tissues. This regulation appears to occur via a posttranscriptional mechanism that involves a direct Ago-dMyc interaction that modulates dMyc stability. dMyc accumulates in ago mutant cells and likely contributes to their increased growth (Moberg, 2004).
The WD repeat domain of Ago interacts with Cyclin E (Moberg, 2001; Strohmaier, 2001; Koepp, 2001), and it also binds dMyc. The optimal binding site for the WD domain of S. cerevisiae Cdc4, the yeast ortholog of Ago, has been determined to be I/L-I/L/P-pT-P-P, in which the central threonine residue is phosphorylated (Nash, 2001). Human Cyclin E, Drosophila Cyclin E, and human c-Myc all have a single, well-conserved version of this site, whose central feature is an L-L-T-P-P motif. dMyc contains seven copies of a degenerate version of this site, in which the central threonine is often replaced by a serine, and many of the flanking residues deviate from those in the consensus sequence. Importantly, these putative sites do retain a conserved S/T residue at position +4. The equivalent +4 serine in human Cyclin E (S384) has been shown to be required for the ubiquitination of Cyclin E (Welcker, 2003) and may therefore represent an important feature of the putative Ago binding motif. The presence of multiple Ago binding sites in dMyc versus the single well-conserved site in c-Myc might indicate that although both proteins are targeted for degradation by orthologous F box proteins, the kinetics of degradation of the two Myc proteins may be different (Moberg, 2004).
The array of apparently suboptimal sites in dMyc resembles the situation in S. cerevisiae Sic1, in which nine low-affinity sites are able to cooperatively mediate a stable interaction with Cdc4 (Welcker, 2003; Orlicky, 2003). Indeed, as is the case with Sic1, mutating a single putative phosphorylation site in dMyc does not alter its Ago binding properties. In contrast, for human Cyclin E and c-Myc, the predicted Ago interaction site lies within a domain previously shown to be required for their ubiquitination and degradation (Salghetti, 1999; Gregory, 2000; Won, 1996; Clurman, 1996). Furthermore, missense mutations of the central threonine in the Ago interaction motif are the most frequent c-Myc mutations in Burkitt's lymphoma and stabilize c-Myc in cells, suggesting that Ago-dependent degradation of c-Myc is perturbed in these cancers (Moberg, 2004).
ago mutant cells grow more quickly than their wild-type neighbors, but they maintain their normal size by an apparent acceleration of the cell cycle. This differs considerably from the phenotype elicited by overexpression of either dMyc or Cyclin E. Increased expression of dMyc results in increased growth that manifests as an increase in cell size without any change in the duration of the cell cycle. dMyc also promotes S phase entry, possibly as a consequence of the increased growth. Increased expression of Cyclin E has no effect on growth but promotes S phase entry. It also results in, at best, a modest acceleration of the cell cycle. Thus, the cell cycle acceleration observed in ago mutant cells is not easily explained by the elevated level of either dMyc or Cyclin E. Both dMyc and Cyclin E promote S phase entry but maintain the normal duration of the cell cycle by apparently lengthening the S and G2 phases, respectively. Thus, it seems likely that ago loss also affects a regulatory protein that promotes the G2-M transition. Such a regulator could either be a direct substrate of SCFAgo or may be regulated indirectly (Moberg, 2004).
Interestingly, both Ago targets identified to date, Cyclin E and dMyc, are required for imaginal-disc growth. Signaling via the Notch receptor is increased in ago clones, as assessed by the activity of a reporter gene fused to the Enhancer of split mβ promoter. Notch signaling has been shown to promote imaginal-disc growth at least in part by a non-cell-autonomous pathway. Because cyclin E, dMyc, and Notch all participate in tissue growth via increases in cell number and/or cell mass, Ago may represent a way to coordinately regulate these pathways by a common degradation mechanism. Thus, increased Ago levels would be expected to impair tissue growth, and decreased levels would facilitate tissue growth, via multiple pathways. Because ago transcription is patterned in the eye imaginal disc (Moberg, 2001), ago may function to link patterning signals with the activity of these growth-promoting pathways (Moberg, 2004).
The ability of ago to regulate multiple pathways that function in growing cells has implications for understanding the role of its human ortholog (Fbw7/hCDC4) as a tumor suppressor gene. Mutations in Fbw7/hCDC4 have been identified in cancer cell lines (Moberg, 2001; Strohmaier, 2001), and more recently, mutations have been identified in Fbw7/hCDC4 in endometrial and colorectal tumors (Spruck, 2002: Rajagopalan, 2004). These tumors are likely to have elevated levels of Cyclin E. In light of the data presented here, they are predicted to have high levels of the oncoprotein c-Myc and increased Notch activity, which has also been implicated in human cancers. Thus, the neoplastic phenotype of these tumors may reflect the additive effect of activating all of these pathways that are normally inhibited by Ago (Moberg, 2004).
Using meiotic and deletion mapping, ago was localized to position 64B on the left arm of chromosome 3. Sequencing candidate transcription units has demonstrated that all three ago alleles (ago1, ago3, ago4) have mutations in an open reading frame designated CG15010. ago encodes a 1,326-amino-acid protein that contains an F-box and seven WD repeats in its carboxy-terminal portion. F-boxes and WD motifs are found in proteins that function as the substrate-recognition component of SCF-type ubiquitin ligase complexes. Within the C-terminal region, the Ago protein is most similar to that encoded by an amino-terminally truncated human expressed sequence tag (EST) previously designated FLJ11071, which has been renamed human Ago. This protein is referred to as hCdc4. A more distant relative is the Caenorhabditis elegans gene sel-10. Two of the mutations are missense mutations that map to the fourth WD repeat. Each alters a residue that is conserved among the three proteins. The third mutation would result in premature termination of translation early in the sixth WD repeat. Since the WD repeats of F-box proteins are thought to be required for recruitment of substrates to the SCF complex, it is possible that the alleles recovered in the screen may impair the interaction of Ago with specific substrates (Moberg, 2002).
date revised: 14 October 2004
Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.