DUB-1A, a Novel Deubiquitinating Enzyme Subfamily Member, Is Polyubiquitinated and Cytokine-inducible in B-lymphocytes*

Recently, we isolated the Dub-2A gene, which encodes a novel murine deubiquitinating enzyme subfamily member, from a bacterial artificial chromosome library clone by PCR amplification with degenerate PCR primers for the Dub-2 cDNA (Baek, K.-H., Mondoux, M. A., Jaster, R., Fire-Levin E., and D'Andrea, A. D. (2001) Blood 98, 636–642). In this study, we analyzed two more clones from the library to isolate genes encoding other deubiquitinating enzymes. Dub-1A, which encodes the shortest member of the DUB subfamily of deubiquitinating enzymes so far, has been identified in both clones and characterized. Sequence analysis showed that Dub-1A encodes a 468-amino acid protein that has a molecular mass of ∼51 kDa and that contains a putative catalytic domain (Cys, His, and Asp) conserved among DUB proteins. The amino acid sequence of DUB-1A is 84.5, 84.7, and 85.3% identical to those of DUB-1, DUB-2, and DUB-2A, respectively. Reverse transcription-PCR revealed that Dub-1A is expressed not only in B-lymphocytes in response to interleukin-3 stimulation, but also in T-lymphocytes, brain, heart, liver, lung, kidney, ovary, and spleen. This suggests that Dub-1A may play essential roles in each of these organs. In vivo and in vitro deubiquitinating enzyme assays showed that DUB-1A has functional deubiquitinating activity and that the 5′-flanking sequence of Dub-1A has a functional enhancer domain as shown in Dub-1 and Dub-2A. Interestingly, immunoblot analysis revealed that DUB-1A is polyubiquitinated, indicating that it is degraded through proteasome-mediated degradation. In the absence of JAK2, Dub-1A was expressed at a lower level. This suggests that DUB-1A functions downstream of JAK2 kinase in the interleukin-3 signaling pathway.

tion with interleukin (IL)-3, IL-5, and granulocyte/macrophage colony-stimulating factor, whereas Dub-2 is induced by stimulation with IL-2 only (2). In addition, it has been reported that DUB-2 is capable of promoting IL-2-mediated signaling and can suppress apoptosis in lymphocytes after withdrawal of growth factor (24,26). However, substrates for these DUB enzymes have not been identified yet, even though it has been suggested that DUB-2 may be involved in reversal of Cbl-bmediated p85 ubiquitination (24,27).
In this study, we describe the cloning of both complementary and genomic DNA for a novel deubiquitinating enzyme, Dub-1A, which is classified as one of the Dub subfamily members. The sequences of the putative polypeptide and enhancer element for Dub-1A are highly homologous to those for previously known Dub subfamily members (Dub-1, Dub-2, and Dub-2A) and reveal deubiquitinating enzyme activity in vivo and in vitro.

Screening of the Bacterial Artificial Chromosome (BAC) Library and
Cloning of the Dub-1A Gene-The murine genomic library in the pBe-loBAC11 vector was screened by genomic PCR with two primers (Bam5Ј, 5Ј-GCGGATCCTTTGAAGAGGTCTTTGGAAA-3Ј; and Xho3Ј, 5Ј-ATCTCGAGGTGTCCACAGGAGCCTGTGT-3Ј) derived from the sequence of the Dub-2 gene. Using the same primers, genomic PCR products containing the Dub-1A gene from two of three positive BAC clones were generated, cloned, and sequenced.
Site-directed Mutagenesis-The Dub-1A(C60S) mutant was generated using a QuikChange TM site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. PCR was performed at 95°C for 30 s, 55°C for 1 min, and 72°C for 14 min for a total of 14 cycles. The forward primer (5Ј-CACAGGCAACAGCAGCTACCTGAAT-GCAGC-3Ј) and reverse primer (5Ј-GCTGCATTCAGGTAGCTGCTGT-TGCCTGTG-3Ј) were used for replacing cysteine with serine at position 60 of DUB-1A.
In Vivo and in Vitro Deubiquitination Assays-The effect of Dub-1A expression on the ubiquitin/proteasome system was analyzed by transfecting pcDNA3-Myc-Dub-1A or pcDNA3-Myc-Dub-1A(C60S) with pMT123-HA-ubiquitin into NIH3T3 cells. After a 24-h transfection, cells were harvested and lysed. 10 g of total proteins were loaded onto each lane of a 10% SDS-polyacrylamide gel for immunoblot analysis using anti-hemagglutinin (HA) antibody (Roche Applied Science). Equal loading was verified by immunoblotting against anti-Myc antibody (Santa Cruz Biotechnology).
In Vitro Isopeptidase Assay-To confirm that DUB-1A has isopeptidase activity, the inhibitory effect of ubiquitin aldehyde on the deubiquitinating activity of DUB-1A was analyzed. For the assay, COS-7 cells were transfected with 10 g of pcDNA3-Myc-Dub-1A using Lipo-fectAMINE according to the manufacturer's protocol. After 36 h, cells were washed with cold phosphate-buffered saline before application of lysis buffer (20 mM Tris (pH 7.5), 1% Triton X-100, 150 mM NaCl, 1 mM EGTA, and 1.5 mM MgCl 2 ) plus protease inhibitor mixture (Roche Applied Science). Lysates were then rotated at 4°C for 1 h followed, by centrifugation at 13,000 rpm for 15 min. The supernatants were immunoprecipitated with anti-Myc antibody and combined with protein A/G-Sepharose (Amersham Biosciences), followed by rotation at 4°C for 4 h. Beads were then washed twice with lysis buffer containing 150 mM NaCl, once using lysis buffer containing 500 mM NaCl, and three times with reaction buffer (50 mM Hepes (pH 7.8), 0.5 mM EDTA, 0.01% Brij, and 3 mM dithiothreitol) before 50% of the beads were used in hydrolysis reactions. To inhibit deubiquitinating activity, DUB-1A was preincubated in the presence of 2 M ubiquitin aldehyde (Affiniti Research Products) for 30 min at 37°C. Mixed polyubiquitin chains (Ub 2-7 ; Affiniti Research Products) were then used as substrates at a final concentration of 1 M as described previously (29). Following hydrolysis, ubiquitins were detected by Western blotting using anti-ubiquitin antibody (Sigma), horseradish peroxidase-conjugated secondary antibody (Zymed Laboratories Inc.), and the ECL detection kit.
In Vivo Co-immunoprecipitation for Ubiquitination Assay-For in vivo co-immunoprecipitation assay, NIH3T3 cells were transfected with 2 g of pcDNA3-Myc-Dub-1A and pMT123-HA-ubiquitin. After a 48-h transfection, co-immunoprecipitation was performed, and precipitates were resolved on a 10% SDS-polyacrylamide gel and subjected to Western blot analysis. Bands were visualized with chemiluminescence (ECL detection kit).
Stable Transfection of Ba/F3 Cells with Wild-type (WT) Jak2 or Dominant-negative (DN)-Jak2-Ba/F3 cells growing in IL-3-containing medium were transfected by electroporation. A cDNA construct (pBOS-WT-Jak2 or pBOS-DN-Jak2) and pSV2neo were applied. G418-resistant Ba/F3 subclones were isolated by limiting dilution in 96-well microtiter plates. Stable expression of WT-Jak2 or DN-Jak2 was confirmed by immunoblot analysis. Each cell line was starved of IL-3 for 8 h and restimulated with 10 pM IL-3 for 2 h. The relative expression level of Dub-1A in transfected Ba/F3 cells was then determined according to the expression level of the glyceraldehyde-3-phosphate dehydrogenase housekeeping gene. The band density of DNA stained with ethidium bromide on a 1% agarose gel was measured using a Gel-Doc image analyzer (Bio-Rad). Average intensities of the DNA band were normalized to those of the glyceraldehyde-3-phosphate dehydrogenase used as a control. Expression levels were tested at least three times by independent transfection.
Isolation of an Enhancer Element and Construction of Luciferase Reporter Plasmids-A 1427-bp fragment corresponding to the promoter region of the Dub-1A gene was amplified by PCR from two BAC clones (BAC2 and BAC3). The primers used for PCR were DUB1e1 (5Ј-CTAG-TAAGGATATAACAGG-3Ј) and T14/CS (5Ј-CATTCAGGTAGCAGCT-GTTGCC-3Ј). The amplified PCR product was subcloned into the pCR2.1-TOPO vector, and the putative enhancer domain (112 bp) from the promoter region was isolated by PCR with primers DUB1e1 and DUB1Ae2 (5Ј-TCTTAGTCACTGTTGTATCT-3Ј) and subcloned into the pGL2 promoter plasmid (Promega), which contains the SV40 basal promoter upstream of the luciferase reporter gene. Mutations of Ets, AP-1, and GATA in the Dub-1A enhancer region generated by PCR using mutant oligonucleotide primers are indicated in Fig. 7A.
Transient Transfection and Transactivation Experiments-Transient transfection of Ba/F3 cells and luciferase reporter gene assays were performed as described previously (30), with minor modifications. Briefly, Ba/F3 cells were washed with phosphate-buffered saline free of serum and IL-3 and cultured in plain RPMI 1640 medium for 4 h. Afterward, they were resuspended at 1 ϫ 10 7 cells/0.8 ml of RPMI 1640 medium and transferred to an electroporation cuvette. The cells were incubated with 10 g of the indicated luciferase reporter vector along with 1 g of a cytomegalovirus promoter-driven ␤-galactosidase reporter gene construct to monitor transfection efficiencies. After electroporation with a Bio-Rad electroporator (350 V, 960 microfarads), the cells were divided into two pools and either restimulated with 10 pM IL-3 for 4 h or left untreated. Luciferase and ␤-galactosidase levels were then analyzed according to manufacturer specifications (luciferase assay kit (Analytical Luminescence Laboratory) and Galacto-Light kit (Tropix), respectively). Each luciferase reporter construct was tested at least three times by independent transfection.

RESULTS
Cloning and Expression of a Novel Dub-1A Gene, the Smallest Dub Subfamily Member-We have previously screened a murine genomic BAC library using Dub-2-specific primers and isolated three BAC clones that may contain a putative genomic Dub-2 gene (25). Unexpectedly, one of these BAC clones (BAC1) contains the Dub-2A gene, which is highly homologous to Dub-2 (25), and the other two BAC clones (BAC2 and BAC3) were subjected to further analysis in this study. Digestion of both BAC clones with different restriction enzymes revealed that they contain the same genomic fragment (data not shown). Sequence analysis of amplified PCR products derived from these clones showed a novel Dub gene (Dub-1A) composed of two exons and one intron (Fig. 1A). This is similar to the structural organization of the genomic DNAs for other Dub subfamily members, including Dub-1, Dub-2, and Dub-2A (2,24,25).
Based on the genomic sequence, the full-length Dub-1A cDNA was subsequently isolated from various mouse tissues by RT-PCR and predicted to contain a 1407-bp open reading frame and to encode a protein of 468 amino acids (Fig. 1B), rendering it the shortest protein among DUB subfamily members (Fig. 2). Exon 1 encodes only 9 amino acids, and exon 2 encodes amino acids 10 -468, divided by 786 bp of the single intron. The TATA box located 107 bp upstream of the start codon is indicated in Fig. 2.
Because the amino acid sequences of DUB subfamily members are highly homologous to one other, we next compared the predicted amino acid sequence of DUB-1A with those previously reported for DUB-1, DUB-2, and DUB-2A (Fig. 2). DUB-1A has 84.5% amino acid identity to DUB-1, 84.7% amino acid identity to DUB-2, and 85.3% amino acid identity to DUB-2A. Like other known ubiquitin-processing protease enzymes, including DUB subfamily members, DUB-1A contains three conserved domains, which are cysteine, histidine, and aspartic acid domains (Fig. 2). Previous reports revealed that these domains are required for deubiquitinating enzyme activity, indicating that they are involved in constructing the active sites of the enzymes (25,31). Interestingly, both DUB-1A and DUB-1 lack a short hypervariable region present in DUB-2 and DUB-2A (amino acids 444 -463, KHR(I/N)NEILPQEQN-HQK(A/T)GQS) (Fig. 2). The cellular function of this region in DUB-2 and DUB-2A has not yet been elucidated. As shown in Fig. 2, the C terminus of DUB-1A is much shorter than those of other DUB subfamily members. These data support the observation that deubiquitinating enzymes function as long as the polypeptide contains those three conserved amino acids (Cys, His, and Asp).
Expression of Dub-1A in Various Tissues and Cytokine-dependent Expression in B-and T-lymphocytes-Due to the presence of Dub-1 in B-lymphocytes, we analyzed the expression pattern of Dub-1A mRNA by RT-PCR using Dub-1A-specific primers (Fig. 1A). We designed specific primers for Dub-1A (1A-S1 and 1A-S2) and demonstrated that they specifically amplified Dub-1A, whereas the degenerate PCR primers recognized all Dub subfamily members (Fig. 3A). Therefore, we used these primers for RT-PCR to determine whether Dub-1A mRNA is expressed in various murine tissues. The sequences of the PCR products showed that Dub-1A is expressed in brain, heart, lung, kidney, ovary, spleen, and B-lymphocytes (Fig.   3B). Interestingly, the expression of Dub-1A in B-lymphocytes was IL-3-dependent. Therefore, we next analyzed the minimal concentration of IL-3 required for Dub-1A to be expressed. After the depletion of IL-3 for 8 h, cells were cultured with different concentrations of IL-3. Dub-1A was not expressed until 10 pM IL-3 or higher was added to the culture medium (Fig. 3C). Dub-1A was constitutively expressed in CTLL-2 cells, indicating that the IL-2 cytokine is not required for the expression of Dub-1A in CTLL-2 cells (Fig. 3D).
Because Dub subfamily members are immediate-early genes, we tested whether Dub-1A is expressed very rapidly in response to cytokine stimulation. Fig. 3D shows that Dub-1A was expressed within 30 min and was maximally expressed in 2 h. This is similar to the expression pattern of other Dub subfamily members, including Dub-1 and Dub-2. After 2 h, the expression of Dub-1A slowly decreased and disappeared after 10 h.
Dub-1A Encodes a Functional Deubiquitinating Enzyme-To determine whether DUB-1A has deubiquitinating enzyme activity, we examined the ability of DUB-1A to cleave ubiquitin from polyubiquitin chains conjugated with proteins via isopeptide bonds under in vivo conditions. HA-tagged ubiquitin was transiently expressed in NIH3T3 fibroblast cells with and without Myc-tagged Dub-1A and its mutant form (Dub-1A(C60S)), followed by immunoblot analysis of the cell extracts using anti-HA antibody (Fig. 4A). The expression of Dub-1A almost completely deubiquitinated polyubiquitinated protein targets in the cells (Fig. 4A, lane 5). However, de- In addition, we expressed DUB-1A as a GST fusion protein to determine whether DUB-1A has deubiquitinating enzyme activity. Immunoblot assays showed that the cDNA clone encoding the GST-DUB-1A fusion protein resulted in cleavage of Ub-Methionine-␤-galactosidase (Fig. 4B, lane 2) to an extent comparable with that observed with GST-DUB-1 (lane 4), GST-DUB-2 (lane 6), and GST-DUB-2A (lane 8). As a control, cells transformed with the pGEX vector failed to cleave Ub-Methionine-␤-galactosidase (lane 1). A mutant form of the DUB-1A polypeptide containing a C60S mutation was unable to cleave the Ub-Methionine-␤-galactosidase substrate (lane 3). These results demonstrate that DUB-1A has deubiquitinating enzyme activity and that cysteine at position 60 is essential for its thiol protease activity, as shown for other DUB subfamily members (2,24,25). These results in vivo and in vitro indicate that DUB-1A has isopeptidase activity.
To confirm that DUB-1A has isopeptidase activity, the effect of ubiquitin aldehyde, a specific inhibitor of deubiquitinating enzymes, on the deubiquitinating activity of DUB-1A was investigated. In the presence of ubiquitin aldehyde, DUB-1A proteins expressed in COS-7 cells were not capable of hydrolyzing the ubiquitin from branched polyubiquitin chains (Fig. 4C), suggesting that DUB-1A has isopeptidase activity.
Ubiquitination of the DUB-1A Protein-Because it has been reported that several proteins, including IB, undergo phosphorylation by kinase before ubiquitination, we therefore analyzed whether DUB-1A is phosphorylated and ubiquitinated or not. Immunoprecipitation assay revealed that the DUB-1A protein was not tyrosine-phosphorylated (data not shown). However, Myc-tagged DUB-1A was coprecipitated with HA-tagged ubiquitin in NIH3T3 cells (Fig. 5, A and B), suggesting that DUB-1A itself is ubiquitinated. It remains to be determined whether the phosphorylation takes place before ubiquitination at either serine or threonine or both amino acids.
JAK2 Signaling Is Required for Induction of Dub-1A Expression-To investigate Dub-1A induction in Ba/F3 cells, we investigated the requirement of JAK2 kinase signaling because Dub-1A was expressed in response to IL-3 stimulation in Blymphocytes (Fig. 3, B and C). Transfection with WT-Jak2 revealed higher expression of Dub-1A (Fig. 6A, lane 2), whereas Dub-1A expression was reduced in Ba/F3 cells transfected with DN-Jak2 lacking the C-terminal tyrosine kinase domain (lane 3), suggesting that DUB-1A functions downstream of JAK2 kinase and requires the presence of WT-JAK2 kinase. This experiment was performed at least three times by independent transfection. This is similar to the expression of Dub-1, which is dependent on JAK2 (30). As shown in Fig. 6B, the expression of Dub-1A in Ba/F3 cells transfected with DN-JAK2 was ϳ30% less compared with that in Ba/F3 cells without transfection.
The Dub-1A Gene Contains a Cytokine-inducible Enhancer Element-We have previously reported that both the Dub-1 and Dub-2A genes contain a cytokine-responsive enhancer element (2,25). It has been demonstrated that the minimal enhancer elements of Dub-1 and Dub-2A are 112 and 100 bp, respectively (25). In an attempt to identify an enhancer region in the Dub-1A gene, we compared the 5Ј-sequences of Dub-1, Dub-2A, and Dub-1A within the enhancer element (Fig. 7A). Interestingly, they are highly homologous to each other and contain Ets, AP-1, and GATA sequences, suggesting the conserved function of the enhancer.
Transcriptional reporter assays in the murine hematopoietic pro-B-lymphocyte cell line (Ba/F3 cells) were performed to test the putative enhancer activity of the Dub-1A region (Fig. 7B). Ba/F3 cells are dependent on murine IL-3 for growth and survival. Ba/F3 cells were transfected with various reporter constructs, including mutant Dub-1A enhancer domains for FIG. 3. Expression analysis of Dub-1A mRNA. A, PCR products generated using a pair of degenerate PCR primers (DUB/CS-down and 4A/T7-2) and a pair of Dub-1A-specific PCR primers (1A-S1 and 1A-S2) were electrophoresed on a 1% agarose gel, which was stained with ethidium bromide. Whereas the degenerate primers recognized all murine DUB enzyme subfamily members, the Dub-1A-specific primers recognized only the Dub-1A sequence. B, the expression of Dub-1A mRNA in Ba/F3 cells (starved, growing, and IL-3-induced cells) and various tissues (brain, heart, liver, lung, kidney, ovary, and spleen) was analyzed by RT-PCR using Dub-1A-specific primers. Dub-1A was expressed in IL-3-induced Ba/F3 cells and in all tissues tested. G3PDH, glyceraldehyde-3-phosphate dehydrogenase. C, Ba/F3 cells were starved and stimulated with various IL-3 concentrations from 0 to 100 pM. Dub-1A was not expressed until 10 pM IL-3 was added to the medium. D, Ba/F3 and CTLL-2 cells were starved and stimulated with murine IL-3 and IL-2, respectively, for the indicated times. Whereas Dub-1A was constitutively expressed in CTLL-2 cells, it was expressed only after induction with the IL-3 cytokine in Ba/F3 cells.
Ets, AP-1, and GATA sequences (Fig. 7A), and IL-3-induced activities were measured (Fig. 7B). As shown in Fig. 7B, the enhancer sequence for Dub-1A had enhancer activity, and mutations at two AP-1 sites and the Ets site in the Dub-1A sequence did not show the activity. This is similar to previous reports that revealed the requirement for AP-1 and Ets sequences in Dub-1 and Dub-2A (25,32). Therefore, it is possible that both Ets and AP-1 (but not GATA) sequences are required for the induction of enhancer activity in Dub subfamily members.

DISCUSSION
It is becoming clear that the regulation of ubiquitination and deubiquitination for protein degradation is essential for cellular processes, including cell proliferation and differentiation (3,20). A number of diseases have been found to be associated with aberrant protein degradation. These include Parkinson's disease, Alzheimer's disease, Angelman syndrome, Huntington's disease, and cystic fibrosis (33), suggesting important roles of protein degradation in normal cellular processes. We have previously reported a hematopoiesis-specific cytokine-inducible gene encoding a growth regulatory deubiquitinating enzyme (DUB-1) (2). Based on the fact that the overexpression of Dub-1 results in cell cycle arrest prior to S phase (2), it is possible that the aberration of Dub expression may lead to abnormal proliferation of lymphocytes, causing leukemia in mammals. Therefore, detailed analysis of deubiquitinating enzymes involved in the regulation of protein degradation is required to better understand their cellular functions.
Recently, three Dub genes have been described and categorized into a novel class of deubiquitinating enzymes: Dub-1 (2), Dub-2 (24), and Dub-2A (25). Both Dub-1 and Dub-2 were identified as hematopoiesis-specific immediate-early genes that are rapidly induced in response to cytokines (2,24). Dub-1 is expressed in B-lymphocytes by stimulation with IL-3, IL-5, and granulocyte/macrophage colony-stimulating factor (2). Dub-2 is expressed in T-lymphocytes by stimulation with IL-2 (24). However, Dub-2A is expressed not only in T-lymphocytes, but also in embryonic stem cells in mouse (25). Interestingly, all DUB enzymes revealed deubiquitinating activity and contain three conserved domains (Cys, His, and Asp) throughout the N terminus. In the absence of these domains, the enzymes are not capable of cleaving ubiquitin, suggesting a critical role in catalytic enzyme reaction. Recently, the structure of the catalytic domain of herpesvirus-associated ubiquitin-specific protease, which deubiquitinates and stabilizes the tumor suppressor p53, has been investigated (34). This investigation revealed that the catalytic core domain of the herpesvirus-associated ubiquitin-specific protease enzyme binds ubiquitin al- dehyde, leading to a dramatic conformational change in the active site similar to the one for herpesvirus-associated ubiquitin-specific protease binding to its substrate (34). Interestingly, the conserved aspartic acid residue participating in the catalytic triad was found downstream of the conserved histidine residues, indicating that the conserved aspartic acid at position 133, localized upstream of the conserved histidine residues in DUB-1 and DUB-2, is involved in the formation of the oxyanion hole instead of catalytic activity. The detailed molecular mechanisms for these enzymes should be investigated to develop pharmaceutical reagents that target these mechanisms in signal transduction pathways involved in a number of cellular processes, including cell proliferation and differentiation.
In this study, we have identified a novel Dub-1A gene from BAC clones that encodes another member of the DUB subfamily of deubiquitinating enzymes. Dub-1A is composed of two exons and one intron and encodes an enzyme highly homologous to other DUB subfamily members (Table I). It has been reported that Dub genes are located in the region of murine chromosome 7 with a tandem repeat array (24). Because both complete Dub-1A and Dub-1 genes are found in the same BAC clones, we expect that Dub-1A is also located in chromosome 7. The sequence similarity among Dub subfamily members and chromosomal co-localization suggest that the Dub genes arose by a tandem duplication of an ancestral Dub gene. Interestingly, Dub-1A is expressed in various tissues, including B-and T-lymphocytes, and encodes the shortest form of DUB protein among subfamily members of deubiquitinating enzymes. DUB-1A, like DUB-1, does not contain the hypervariable domain that is present in both DUB-2 and DUB-2A (Fig. 2). Because the C terminus of Dub-1A tends to encode different amino acids compared with other DUB subfamily members, we propose that the C terminus rather than the N terminus has substrate specificity. This supports the previous observation that deubiquitinating  activity is retained upon deletion of the C terminus as long as the three conserved domains are intact (25,31).
It has been reported that JAK2 kinase and the Ras-Raf-MEK-ERK kinase signaling pathway are required for induction of the murine Dub-1 gene, even though the presence of another signaling pathway may be involved (30). Interestingly, the minimal interleukin-3-responsive element of the Dub-1 gene contains cytokine-inducible enhancer activity, but lacks a consensus sequence for STAT binding. This indicates that Dub-1 is expressed in a JAK2-dependent, but STAT5-independent pathway (30). In the case of the Dub-1A enhancer sequence, the consensus sequence for STAT binding is also not present. However, blocking JAK2 signaling inhibits the expression of Dub-1A (Fig. 6, A and B), indicating the requirement of JAK2 signaling for Dub-1A to be expressed. Because many receptor tyrosine kinases are regulated not only by phosphorylation and dephosphorylation, but also by ubiquitination and deubiquitination, it is possible that signaling pathways mediated by these receptors can be modulated by the ubiquitination status of them. Therefore, finding the molecular mechanisms for the regulation of protein degradation via ubiquitination and deubiquitination in receptor tyrosine kinase-mediated signal transduction pathways will contribute to understanding the regulation of cell proliferation and differentiation. Interestingly, DUB-1A itself is ubiquitinated (Fig. 5, A and B), suggesting that this protein is also regulated by proteasomemediated degradation when it is no longer necessary within cells.
Because Dub-1 and Dub-2A contain a cytokine-inducible enhancer element, we have further characterized the enhancer element of Dub-1A. Transcriptional reporter assays in Ba/F3 cells revealed that two AP-1 sites and one Ets site are required for Dub-1A enhancer activity. This is similar to Dub-1 and Dub-2A (25,31), suggesting that they play a key role in responding to cytokine stimuli. Even though substrates for each DUB enzyme remain to be found, it will be helpful to identify ways of regulating their expression. This will give us insights into the regulation of lymphocyte proliferation and immune responses in vivo. It has been suggested that Dub-2 induced by IL-2 might be involved in the regulation of T cell receptor clustering for the supramolecular activation complex by deubiquitinating p85, the phosphatidylinositol 3-kinase adaptor subunit (27). The functional roles of DUB enzymes in signal transduction in immunity have to be investigated to provide new possibilities for designing therapeutic drugs.
Acknowledgments-We thank Alan D. D'Andrea (Harvard Medical School) and members of the Cell and Gene Therapy Research Institute of the Pochon CHA University for critical comments on the manuscript.