A dominant insertional P-element mutation enhances position-effect variegation in Drosophila melanogaster. The mutation is homozygous, viable, and fertile and maps at 64E on the third chromosome. The corresponding gene was cloned by transposon tagging. Insertion of the transposon upstream of the open reading frame correlates with a strong reduction of transcript level. A transgene was constructed with the cDNA and found to have the opposite effect from that of the mutation, namely, the suppression of variegation. Sequencing of the cDNA reveals a large open reading frame encoding a putative ubiquitin-specific protease (Ubp). Ubiquitin marks various proteins, frequently for proteasome-dependent degradation. Ubps can cleave the ubiquitin part from these proteins. The deduced protein, termed D-Udp-64E, consists of 898 amino acids and shares a Cys and His domain structure with other ubiquitin C-terminal hydrolases. The transcript is expressed ubiquitously during embryonic development (Henchoz, 1996).

In eukaryotes, both natural and engineered ubiquitin (Ub) fusions, either to itself or to other proteins, are cleaved by processing proteases after the last amino acid residue of ubiquitin. YUH1 and UBP1, the genes for two ubiquitin-specific proteases of the yeast Saccharomyces cerevisiae, have been cloned previously and shown to encode nonhomologous proteins. Using an Escherichia coli-based genetic screen, two other yeast genes for ubiquitin-specific proteases have been isolated and named UBP2 and UBP3. Ubp2 (1,264 residues), Ubp3 (912 residues), and the previously cloned Ubp1 (809 residues) are largely dissimilar except for two short regions containing Cys and His which encompass their putative active sites. Neither of these proteases has sequence similarities to Yuh1. Both Ubp2 and the previously identified Ubp1 cleave in vitro at the C terminus of the ubiquitin moiety in natural and engineered fusions irrespective of their size, poly-Ub being the exception. However, both Ubp1 and Ubp2 are also capable of cleaving poly-Ub when coexpressed with it in E. coli, suggesting that such cleavage is largely cotranslational. Although inactive in E. coli extracts, Ubp3 is active with all of the tested ubiquitin fusions except poly-Ub when coexpressed with them in E. coli. Null yuh1 ubp1 ubp2 ubp3 quadruple mutants are viable and retain the ability to deubiquitinate ubiquitin fusions, indicating the presence of at least one more ubiquitin-specific processing protease in S. cerevisiae (Baker, 1992).

A necessary step in ubiquitin-dependent proteolysis is the addition of a polyubiquitin chain to the target protein. This ubiquitinated protein is degraded by a multisubunit complex known as the 26S proteasome. The polyubiquitin chain is probably not released until a late stage in the proteolysis by the proteasome. It is subsequently disassembled to yield functional ubiquitin monomers. A 93 kDa protein, isopeptidase T, has the properties expected for the enzymethatdisassembles these branched polyubiquitin chains. Protein and cDNA sequencing reveals that isopeptidase T is a member of the ubiquitin specific protease family (UBP). Isopeptidase T is a protein of 835 amino acids with conserved Cys and His boxes characteristic of UBPs. Isopeptidase T disassembles branched polyubiquitin chains (linked by the G76-K48 isopeptide bond) by a sequential exo mechanism, starting at the proximal end of the chain (the proximal ubiquitin contains a free carboxyl-terminus). Isopeptidase T prefers to disassemble chains in which there is an intact and unblocked RGG sequence at the C-terminus of the proximal subunit. Rates of disassembly are reduced when G76 of the proximal ubiquitin is modified, for example, by ligation to substrate protein, by esterification, by replacement of the proximal glycine with alanine (G76A), or by truncation. Linear proubiquitin provides only a poor substrate. Observed rates and specificity are consistent with isopeptidase T playing a major role in disassembly of polyubiquitin chains. The high discrimination against chains that are blocked or modified at the proximal end indicates that the enzyme acts after release of the chains from conjugated proteins or degradation intermediates. Thus, the proteolytic degradation signal is not disassembled by isopeptidase T before the ubiquitinated protein is degraded. These (and earlier) results suggest that UBP isozymes may exhibit significant substrate specificity, consistent with a role in the regulated catabolism of the polymeric ubiquitin, including the polyubiquitin protein degradation signal (Wilkinson, 1995).

Using a gene trap approach in embryonic stem cells, a murine gene has been isolated with extensive sequence similarity to the Drosophila faf gene, and been termed Fam (fat facets in mouse). The putative mouse protein shows colinearity and a high degree of sequence identity to the Drosophila protein over almost its entire length of 2554 amino acids. The two enzymatic sites characteristic of ubiquitin-specific proteases are very highly conserved between mice and Drosophila and this conservation extends to yeast. Fam is expressed in a complex pattern during postimplantation development. In situ hybridization detects Fam transcripts in the rapidly expanding cell populations of gastrulating and neurulating embryos, in post-mitotic cells of the CNS as well as in the apoptotic regions between the digits, indicating that it is not associated with a single developmental or cellular event. The strong sequence similarity to faf and the developmentally regulated expression pattern suggest that Fam and the ubiquitin pathway may play a role in determining cell fate in mammals, as has been established for Drosophila (Wood, 1997).

EST 221 derived from human adult testis detects homology to the Drosophila fat facets gene (fat) and has related sequences on both the X and Y chromosomes mapping to Xp11.4 and Yq11.2, respectively. These two loci have been termed DFFRX and DFFRY for Drosophila fat facets related X and Y. The major transcript detected by EST 221 is about 8 kb in size and is expressed widely in a range of 16 human adult tissues. RT-PCR analysis of 13 different human embryonic tissues, with primers specific for the X and Y sequences demonstrates that both loci are expressed in developing tissues. Quantitative RT-PCR analysis of lymphoblastoid cell lines carrying different numbers of X chromosomes reveals that the X-linked gene escapes X-inactivation. The amino acid sequence (2547 residues) for the complete open reading frame of the X gene has 44% identity and 88% similarity to the Drosophila sequence and contains the conserved Cys and His domains characteristic of deubiquitinating enzymes, suggesting its biochemical function may be the hydrolysis of ubiquitin from protein-ubiquitin conjugates. The requirement of faf for normal oocyte development in Drosophila combined with the map location and escape from X-inactivation of DFFRX raises the possibility that the human homolog plays a role in the defects of oocyte proliferation and subsequent gonadal degeneration found in Turner's syndrome (Jones, 1996).

Human DFFRY (the Y-linked homologue of the DFFRX Drosophila fat-facets related X gene) maps to proximal Yq11.2 within the interval defining the AZFa spermatogenic phenotype. The complete coding region of DFFRY has been sequenced and shows 89% identity to the X-linked gene at the nucleotide level. In common with DFFRX , the potential amino acid sequence contains the conserved Cys and His domains characteristic of ubiquitin C-terminal hydrolases. The human DFFRY mRNA is expressed in a wide range of adult and embryonic tissues, including testis, whereas the homologous mouse Dffry gene is expressed specifically in the testis. Analysis of three azoospermic male patients has shown that DFFRY is deleted from the Y chromosome in these individuals. Two patients have a testicular phenotype, which resembles Sertoli cell-only syndrome, and the third shows diminished spermatogenesis. In all three patients, the deletions extend from close to the 3' end into the gene, removing the entire coding sequence of DFFRY . The mouse Dffry gene maps to the Sxrb deletion interval on the short arm of the mouse Y chromosome; its expression in mouse testis can first be detected between 7.5 and 10. 5 days after birth when type A and B spermatogonia and pre-leptotene and leptotene spermatocytes are present (Brown, 1998).

A cDNA encoding a new ubiquitin-specific protease, UBP41, in chick skeletal muscle was cloned using an Escherichia coli-based in vivo screening method. Nucleotide sequence analysis of the cDNA containing an open reading frame of 1,071 base pairs reveals that the protease consists of 357 residues with a calculated molecular mass of 40,847 Da, and that it is related to members of the UBP family containing highly conserved Cys and His domains. Chick UBP41 was expressed in E. coli and purified from the cells to apparent homogeneity. The purified enzyme behaves as an approximately 43-kDa protein under both denaturing and nondenaturing conditions, suggesting that it consists of a single polypeptide chain. Like other deubiquitinating enzymes, it was sensitive to inhibition by ubiquitin-aldehyde and sulfhydryl blocking agents, such as N-ethylmaleimide. The UBP41 protease cleaves at the C terminus of the ubiquitin moiety; thus, the protease is active against ubiquitin-beta-galactosidase as well as ubiquitin C-terminal extension protein of 80 amino acids. UBP41 also releases free ubiquitin from poly-His-tagged di-ubiquitin. It converts poly-ubiquitinated lysozyme conjugates to mono-ubiquitinated forms of about 24 kDa, although the latter molecules are not further degraded to free ubiquitin and lysozyme. These results suggest that UBP41 may play an important role in the recycling of ubiquitin by hydrolysis of branched poly-ubiquitin chains generated by the action of 26 S proteasome on poly-ubiquitinated protein substrates, as well as in the production of free ubiquitin from linear poly-ubiquitin chains and of certain ribosomal proteins from ubiquitin fusion proteins (Baek, 1997).

Cytokines regulate cell growth by inducing the expression of specific target genes. Using the differential display method, cytokine-inducible immediate early gene, DUB-1 (for deubiquitinating enzyme), has been cloned. DUB-1 is related to members of the UBP superfamily of deubiquitinating enzymes that includes the oncoprotein Tre-2. A glutathione S-transferase-DUB-1 fusion protein cleaves ubiquitin from a ubiquitin-beta-galactosidase protein. When a conserved cysteine residue of DUB-1, required for ubiquitin-specific thiol protease activity, is mutated to serine (C60S), deubiquitinating activity is abolished. Continuous expression of DUB-1 from a steroid-inducible promoter induces growth arrest in the G1 phase of the cell cycle. Cells arrested by DUB-1 expression remain viable and resume proliferation upon steroid withdrawal. These results suggest that DUB-1 regulates cellular growth by modulating either the ubiquitin-dependent proteolysis or the ubiquitination state of an unknown growth regulatory factor(s) (Zhu, 1996).

The Ras target AF-6 has been shown to serve as one of the peripheral components of cell-cell adhesions, and is thought to participate in cell-cell adhesion regulation downstream of Ras. An AF-6-interacting protein with a molecular mass of approximately 220 kD (p220) waa purified to investigate the function of AF-6 at cell-cell adhesions. The peptide sequences of p220 are identical to the amino acid sequences of mouse Fam. Fam is homologous to a deubiquitinating enzyme in Drosophila, the product of the fat facets gene. Recent genetic analyses indicate that the deubiquitinating activity of the fat facets product plays a critical role in controlling the cell fate. Fam accumulates at the cell-cell contact sites of MDCKII cells, but not at free ends of plasma membranes. Fam is partially colocalized with AF-6 and interacts with AF-6 in vivo and in vitro. AF-6 is ubiquitinated in intact cells, and Fam prevents the ubiquitination of AF-6 (Taya, 1998).

In the ubiquitin-proteasome pathway, the ubiquitinated substrates either undergo degradation by the proteasome or stabilization through the action of the deubiquitinating enzyme. The deubiquitinating enzyme Fam is colocalizes with AF-6, one of the effectors of the Ras small GTPase, at cell-cell contact sites in epithelial cells and interacts with AF-6 in vivo and in vitro. Fam has deubiquitinating activity in vitro and prevents the ubiquitination of AF-6 in intact cells. The degradation of beta-catenin, which accumulates at the cell-cell contact sites as a cadherin/catenin complex, is thought to be regulated by the ubiquitin-proteasome pathway. These observations prompted an examination of the possible Fam regulation of the stabilization of beta-catenin. It was found that Fam interacts with beta-catenin both in vivo and in vitro. The Fam-binding site of beta-catenin maps to the region close to the APC or Axin-binding site of beta-catenin. Over-expression of Fam in mouse L cells results in an elevation of beta-catenin levels and in an elongation of the half-life of beta-catenin. In these L cells, Fam is colocalized with beta-catenin at the dot-like structures in the cytoplasm. These results indicate that Fam interacts with and stabilizes beta-catenin in vivo, presumably through the deubiquitination of beta-catenin (Taya, 1999).

The Drosophila fat facets and canoe genes regulate non-neural cell fate decisions during ommatidium formation. The FAM (Fat facets in mouse) de-ubiquitinating enzyme regulates the function of AF-6 (mammalian Canoe homolog) in the MDCK epithelial cell line. The expression of the FAM and AF-6 proteins overlaps extensively in the mouse eye from embryogenesis to maturity, especially in the non-neural epithelia including the retinal pigment epithelium, subcapsular epithelium of the lens and corneal epithelium. Expression is not limited to the epithelia however, because FAM and AF-6 also co-localize during lens fiber development as well as in sub-populations of the neural retina (Kanai-Azuma, 2000).

Fat facets is a Drosophila deubiquitinating enzyme required for eye development and early embryogenesis. Genetic evidence suggests that Fat facets deubiquitinates and thereby prevents the proteasomal degradation of specific substrates. The Drosophila Liquid facets protein is implicated as the critical substrate of Fat facets in the eye. A mouse homolog of Fat facets, called Fam, has been identified. The results of biochemical experiments implicate two different proteins, Af-6 and beta-catenin, as substrates for Fam. The functional relationship between Fat facets and Fam has been explored. Fam can substitute for Fat facets in all of its essential functions in Drosophila. In addition, the hypothesis was tested that Canoe and Armadillo, the Drosophila homologs of Af-6 and beta-catenin, respectively, are important substrates for Fat facets in the Drosophila eye. No genetic evidence has been found to support a role for either Canoe or Armadillo in the essential Fat facets pathways in Drosophila eye development. The significance of these results is discussed in light of the biochemical experiments that suggest that Af-6 and beta-catenin are substrates of Fam (Chen, 2000a).

While Faf is similar to other Ubps mainly in the small Cys and His domains that define the catalytic region, Faf and Fam share approximately 50% identity and 70% similarity in amino acid sequence along the entire length of the two proteins. The high degree of primary sequence conservation in the two proteins suggests that Faf and Fam functions might be conserved. Since Fam-expressing transgenes can substitute for the endogenous faf gene in the eye and the ovary, the only two tissues in which faf function is critical in Drosophila, it is concluded that the Faf and Fam protein functions have indeed been conserved. Although yeast Ubp2 and Ubp3 bear minimal primary amino acid sequence similarity with Faf outside the catalytic domain, the Ubp2 and Ubp3 genes also can substitute to some extent for faf in the Drosophila eye. These observations do not weaken the present argument that Faf and Fam are functionally homologous for two reasons: (1) neither Ubp2 nor Ubp3 transgenes complement the faf mutant eye phenotype nearly as well as do faf or Fam transgenes; (2) Ubp2 expression in the Drosophila eye results in a mutant phenotype; this was never observed with any faf or Fam transgenes. Thus, faf and Fam behave identically when expressed in the Drosophila eye, while Ubp2 and Ubp3 behave differently from each other and from faf and Fam (Chen, 2000a).

The genetic interaction experiments presented in this study suggest that Cno and Arm, two cell-cell junction proteins thought to be important for integrating signaling with cell adhesion, and whose vertebrate homologs (Af-6 and beta-Catenin) are implicated as substrates of Fam, are not likely to be important substrates of Faf in the eye. In addition, the results of previous experiments implicate the Lqf protein, a Drosophila homolog of the vertebrate endocytosis complex component epsin, as the one critical substrate of Faf in the eye (Cadavid, 2000). If beta-Catenin and Af-6 are in vivo substrates of Fam, how can the failure to detect genetic interactions between arm or cno and faf be explained? One possibility is that Faf could regulate Arm and/or Cno in the eye, but that this role for Faf is non-essential in Drosophila. In fact, faf has a normally redundant role in Ras1 signal transduction in the eye. This idea was tested genetically for cno in the eye and no evidence of a role for faf in a cno pathway was revealed (Chen, 2000a).

A second possibility is that Arm and/or Cno is the essential substrate of Faf in the ovary. Embryos from females homozygous for faf mutations die very early during development; many rounds of nuclear division take place and some abnormally clustered pole cells (primordial germ cells) form but somatic cellularization never occurs. It is unknown whether cno or lqf are expressed or required in the ovary. The arm gene, however, is normally expressed in the female germline where it is required for a multitude of cell adhesion functions. The failure of cellularization in faf mutant embryos and the improper clustering of the pole cells suggest the possibility of a role for Arm in this pathway. Further experiments are required to determine if any of the three proteins are critical substrates for Faf in the Drosophila ovary (Chen, 2000a).

If Arm and Cno are true substrates for Fam in the mouse, it is unknown whether their interaction is essential for mouse development. The Fam gene is expressed widely in the mouse but Fam function may not be essential in all of these tissues. Indeed, Drosophila faf is expressed in many fly tissues (male sex organs, gut, fat body) where it is non-essential. A third possibility is that Arm/beta-catenin and Cno/Af-6 are not substrates of Faf/Fam in vivo. The catalytic domain of Fam alone was shown to bind to Af-6. Since the catalytic domains of Ubps are conserved, it is possible that, in vivo, another Ubp normally binds to Cno and/or Af-6. It has not been reported whether other Ubp catalytic domains can bind Cno and/or Af-6. Recently, it has been shown that the ENTH (epsin N-terminal homology) domain of epsin/Lqf bears structural similarity with the Armadillo repeats present in Arm/beta-catenin. Since beta-catenin binds Fam through these repeats (Taya, 1999), it is possible that Fam/beta-catenin binding is an artifact of the similarity between Lqf and beta-catenin. Alternatively, the structural similarity between the ENTH domain and the Armadillo repeats may indicate that the Faf/Fam substrates constitute a family of proteins containing this structural domain. The region of vertebrate Af-6 that binds Fam is not conserved at the primary amino acid level with its Drosophila homolog, Cno. Perhaps there is some underlying structural similarity between Af-6 and Cno in this region, which could also be shared by the ENTH domain and Arm repeats, and which is not apparent at the primary amino acid sequence level. These speculations highlight appealing avenues for further experiments (Chen, 2000a).

During the Drosophila oogenic processes, Fat facets (Faf), an ubiquitin-specific protease essential for normal development of oocyte and eye, becomes localized at the posterior pole and is incorporated into the pole cells. This is dependent on Oskar, a key factor for pole cell determination, and suggests a role for Faf in germ cell differentiation and development. This study shows that Usp9x, an X-linked ortholog of Faf, is predominantly expressed in both germ cell and supporting cell lineages during mouse gonadal development in stage- and sex-dependent manners. Usp9x was first detected in PGCs at 10.5 days post coitum (dpc), and thereafter its expression both at mRNA and protein levels is enhanced in PGCs of both sexes at 11.5-13.5 dpc. In testis, Usp9x expression rapidly decreases to an undetectable level by 15.5 dpc and after birth to adult, no expression is found in any spermatogenic cells, except for weak expression in Sertoli cells. In the ovary, Usp9x expression in embryonic oocytes is also reduced at the newborn stage, its expression reappears in oocytes at the secondary follicle stage, and its products are highly accumulated in the cytoplasm of Graaffian follicles in adults. Although follicular epithelial cells also express Usp9x at a moderate level during postnatal development, its expression is downregulated from early secondary follicle stage. Thus, the present study is not only the first to demonstrate a conserved expression of fat facets in PGCs between mouse and fly, but also sex- and stage-dependent changes of a specific component of the deubiquitylation system during mammalian gonadal development (Noma, 2002).