Identification of a dominant apoptosis-inducing factor which is involved in the ubiquitin/proteasome dependent pathways of protein degradation.
Andreas Gewies, Ulla Cramer and Stefan Grimm, Junior Research Group for Programmed Cell Death at the Max-Planck-Institute for Biochemistry, Am Klopferspitz 18a, 82152 Martinsried, Germany.
email: andreas@gewies.com
logo
Results
By applying a direct functional genetic screen for dominant apoptosis-inducing genes using a normalized mouse kidney cDNA library (Grimm and Leder, 1997, J. Exp. Med., 185(6): 1137-1142), we identified several cDNAs which upon overexpression induce cell death in the transformed human kidney cell line 293T. One of those cDNA clones was designated UI64, which not only induced the typical apoptotic morphological changes but also produced oligonucleosomal DNA fragmentation as a hallmark event of apoptosis. Sequence analysis identified UI64 with the mouse homologue (gi:7949157) of ubiquitin-specific protease UBP41 (also denominated USP2) which recently had been isolated and characterized in chicken skeletal muscle (Baek et al, 1997; JBC, 272(41): 25560-65). Therefore, the human homologue of UBP41, subsequently designated hUBP41, was isolated by RT-PCR based on its Genbank entry (gi: 4759291) and its coding sequence was cloned adding a C-terminal HA tag. Site-directed mutagenesis of the active-site cysteine to alanine delivered the enzymatically inactive form hUBP41 C24A HA. Transfection of hUBP41 HA and hUBP41 C24A HA into 293T cells revealed that human UBP41 induces apoptosis beginning after about 42h as judged by phenotype and DNA laddering, whereas the active site mutant did not induce significant cell death. Additionally, the two homologous USP family members hUSP18 (gi:8394518, 25% identity to hUPB41) and hUSP21 (gi:7706754, 45% identity to hUBP41) were cloned as C-terminally HA tagged proteins as well as their corresponding active site mutants hUSP18 C64A HA, and hUSP21 C37A HA. Transient transfection into 293T cells demonstrated that only overexpression of hUBP41 HA resulted in significant DNA fragmentation and caspase-3 activity as well as apoptotic phenotype. When transfected into HeLa cells, hUBP41 again clearly induced apoptosis 30h post transfection, although the active-site mutant hUBP41 C24A also induced significant levels of cell death in this cell system. In HeLa cells, DNA fragmentation induced by hUBP41 could be inhibited by the pan-caspase inhibitor zVAD-fmk and by co-transfection with Bcl-XL, supporting an apoptotic mode of cell death. Anti-Ubiquitin immunoblot analysis against total lysates from 293T cells transfected with UBP41 HA showed the disappearance of a broad range of ubiquitinated protein bands when compared to extracts from control transfected cells or cells transfected with the inactive-site mutant hUBP41 C24A HA. Thus, UBP41 HA apparently is able to deubiquitinate a broad range of proteins when expressed at high levels, whereas its homologues hUSP18 HA and hUSP21 HA are not that efficient, although the ubiquitin-signal in hUSP21 HA transfected 293T also appeared to be slightly diminished. The same deubiquitinating activity of hUBP41 HA could be observed in HeLa cells. We asked whether the deubiquitinating activity of hUBP41 HA can result in the stabilization of an artificial N-terminal ubiquitinated and therefore unstable test substrate, the ubiquitin-fusion protein Ub-G76V-GFP (Dantuma et al., 2000, Nat. Biotech., 18: 538-543). In HeLa cells, hUBP41 HA is able to stabilize this ubiquitinated GFP fusion protein, but unexpectedly also the enzymatically active-site mutant hUBP41 C24A HA could stabilize this Ub-G76V-GFP substrate. The same effect could be observed in 293T cells. Fluorescence microscopy of EYFP fusion constructs of hUBP41 and hUBP41 C24A transfected into HeLa cells suggest an even distribution of hUBP41 EYFP being present within the cytoplasm as well as in the nucleus. hUBP41 C24A EYFP also was evenly distributed all over the cell but in addition usually displayed spots of higher fluorescence intensity, possibly representing protein aggregates.

Discussion
As part of the ubiquitin-proteasome system, deubiquitinating enzymes have been demonstrated to be involved in a variety of biological processes such as development, growth, and transcription. Although there is accumulating evidence for an involvement of the ubiquitin proteasome system in apoptosis, thus far there have been only few reports implicating DUBs in the apoptotic process. One example is the fat facets gene (faf) in Drosophila, which encodes a ubiquitin-specific protease apparently involved in the regulation of cell number during development of the fly eye. Moreover, Fam (a mouse homologue of faf) was detected to be expressed in the apoptotic regions between the digit condensations of the developing limbs (Wood et al., 1997, 63: 29-38). It also has been reported that an active-site mutant of Ubp-M eventually leads to apoptosis induction when transfected into mammalian cells (Cai et al., 1999, PNAS, 96:2828-33). Here we show that overexpression of the ubiquitin-specific protease hUBP41 (also designated USP2) induces cell death in human 293T and HeLa cells. The apoptotic mode of cell death is supported by the detection of oligonucleosomal DNA fragmentation, activation of caspase-3, and by the fact that cell death in HeLa can be inhibited by the caspase inhibitor zVAD-fmk and upon co-expression with the anti-apoptotic Bcl-2 family member Bcl-XL. Overexpression of the homologous members hUSP18 or hUSP21 in 293T cells did not result in apoptosis induction and it cannot be excluded that this is explained by the fact that expression levels of hUSP18 and hUSP21 were significantly lower than those of hUBP41. In 293T cells, only wildtype hUBP41 was able to induce cell death whereas its active-site mutant hUBP41 C24A was not, suggesting that the enzymatic activity of hUBP41 was necessary for its apoptotic effect. Unexpectedly, in HeLa cells also the active-site mutant hUBP41 C24A induced significant cell death, and it also was surprising that this mutant could stabilize the ubiquitinated GFP test substrate Ub-G76V-GFP to the same extend as the wildtype enzyme hUBP41. Possibly, the mutant enzyme can still bind to this ubiquitinated substrate, but since not being able to deubiquitinate, it keeps the substrate bound in a complex in which the ubiquitin moiety is shielded from recognition by the proteasome. When EYFP fusions of hUBP41 and its active-site mutant were observed under the fluorescence microscope it could be seen that in contrast to wildtype hUBP41, the mutant enzyme obviously formed aggregates within the cell, which might contribute to hUBP41 C24A induced cell death. The enzymatically inactive mutant of Ubp-M was described to tightly associate with mitotic chromosomes, thereby possibly mediating its apoptotic effect upon overexpression (Cai et al., 1999, PNAS, 96:2828-33). When hUBP41 is overexpressed in 293T and HeLa cells, it strongly deubiquitinates a broad range of proteins and thus should strongly interfere with the ubiquitin-proteasome system. Interestingly, a similar diminution of overall protein ubiquitination and concurrent apoptosis induction was described in Balb/c 3T3 fibroblasts with a defect in the E1 ubiquitin-activating enzyme (Monney et al., 1998, JBC, 273(11): 6121-6131). Monney et al. proposed that the decreased ubiquitination possibly causes accumulation of otherwise short-lived proteins (such as p53, Myc, Jun, E2F,) which subsequently trigger apoptosis. Apoptosis induced by overexpression of hUBP41 also might be caused by the accumulation of pro-apoptotic factors, and this possibility will be one subject of future investigations.