Blockade of the ubiquitin protease UBP43 destabilizes transcription factor PML/RARα and inhibits the growth of acute promyelocytic leukemia

Abstract

More effective treatments for acute promyelocytic leukemia (APL) are needed. APL cell treatment with all-trans-retinoic acid (RA) degrades the chimeric, dominant-negative-acting transcription factor promyelocytic leukemia gene (PML)/RARα, which is generated in APL by chromosomal translocation. The E1-like ubiquitin-activating enzyme (UBE1L) associates with interferon-stimulated gene ISG15 that binds and represses PML/RARα protein. Ubiquitin protease UBP43/USP18 removes ISG15 from conjugated proteins. In this study, we explored how RA regulates UBP43 expression and the effects of UBP43 on PML/RARα stability and APL growth, apoptosis, or differentiation. RA treatment induced UBE1L, ISG15, and UBP43 expression in RA-sensitive but not RA-resistant APL cells. Similar in vivo findings were obtained in a transgenic mouse model of transplantable APL, and in the RA response of leukemic cells harvested directly from APL patients. UBP43 knockdown repressed PML/RARα protein levels and inhibited RA-sensitive or RA-resistant cell growth by destabilizing the PML domain of PML/RARα. This inhibitory effect promoted apoptosis but did not affect the RA differentiation response in these APL cells. In contrast, elevation of UBP43 expression stabilized PML/RARα protein and inhibited apoptosis. Taken together, our findings define the ubiquitin protease UBP43 as a novel candidate drug target for APL treatment.

Fig. 1

All-trans-retinoic acid (RA) regulation of UBE1L, ISG15, and UBP43 mRNA expression in RA-sensitive versus RA-resistant acute promyelocytic leukemia (APL) cells. (A) RA-sensitive NB4-S1 cells and RA-resistant NB4-R1 APL cells were each cultured with RA (1μM) or with vehicle for the indicated times in hours (h). UBE1L, ISG15, UBP43 and GAPDH mRNAs were each assessed by real-time RT-PCR assays. Similar results were obtained from three independent experiments. A representative result is shown. RA-treatment induced UBE1L and ISG15 before UBP43 expression in NB4-S1 cells, but these species were not induced in NB4-R1 APL cells. (B and C) Immunogenic peptides generated the indicated anti-UBP43 polyclonal antibodies (see Materials and Methods). UBP43 protein was detected in COS-7 cells transfected with pcDNA4-UBP43 (+), but not with an insertless control pcDNA4 vector (-). The 43-kDa UBP43 protein was detected using these respective antibodies. This protein was not detected when pre-immune antisera for each antibody was probed to each respective filter. Molecular weight size markers are displayed.

Fig. 2

UBP43 is expressed and active in APL cells. (A) ISG15-UBP43 conjugation was detected in NB4-S1 APL cells. Immunoprecipitation (IP) with an anti-HA antibody followed by immunoblotting (IB) with a second anti-HA antibody (left panel) or an anti-UBP43 antibody (middle panel) revealed ISG15-UBP43 complex formation in NB4-S1 APL cells pre-incubated with HA-ISG15-vinylmethyl ester (HA-ISG15-VME). Closed arrows depict positions of UBP43 conjugated with ISG15, the open arrow indicates unconjugated UBP43 protein, and the hatched arrow shows the position of HA-tagged ISG15. The absence of HA-tagged ISG15 detection in the first lane of the left panel of Fig. 2A is due to no HA-tagged ISG15 addition to this lysate. Other species detected in the left and middle panels likely represent non-specific signals. The right panel indicates that UBP43 can remove ISG15 from PML/RARα protein following anti-HA immunoprecipitation of HA-tagged PML/RARα protein followed by anti-HA or anti-ISG15 immunoblotting (IB) to detect PML/RARα. Molecular weight size markers are displayed. (B and C) Independent immunoblot expression for UBP43 in NB4-R1 and in NB4-S1 APL cells treated with RA (1μM) or vehicle (DMSO) for the indicated hours (h). No appreciable change in UBP43 expression occurred after RA-treatment of NB4-R1 APL cells, as shown in panel B. In contrast, UBP43 was induced after 48 hours of RA-treatment in NB4-S1 APL cells, as displayed in panel C. Quantifications of respective signals are presented.

Fig. 3

Effects of UBP43 transfection on PML/RARα protein stability. (A) Three different constructs, designated 1 through 3, assessed PML/RARα protein domain stabilities after engineered UBP43 over-expression. The EGFP expression vector served as a transfection control. Actin expression confirmed similar amounts of loaded proteins. Quantifications of these indicated immunoblot signals appear in the right panels. Full-length PML/RARα and the PML, but not RARα domains of PML/RARα were stabilized by UBP43. (B) Effects of UBP43 on transfected HA-tagged PML/RARα in BEAS-2B cells. Immunoblot expression for PML/RARα and actin is shown in the presence (+) or absence (−) of cycloheximide (CHX; 40μg/mL) treatment. UBP43 stabilized exogenous PML/RARα protein despite CHX-treatment, as detected by the anti-HA antibody. Quantification of signals is provided. (C and D) UBP43 knock-down in APL cells. (C) Transient transfection of two independent UBP43 targeting siRNAs versus an inactive control siRNA (see Materials and Methods) reduced UBP43 and decreased PML/RARα immunoblot expression in NB4-S1 cells, as in panel C. (D) UBP43 knock-down by each siRNA targeting UBP43 significantly (*, P < 0.01) induced apoptosis versus inactive control siRNA in panel D.

Fig. 4

(A) Stable UBP43 knock-down by retroviral transduction of UBP43 shRNA in APL cells. UBP43 shRNA transduction (+) destabilized endogenous PML/RARα protein expression, as compared to an insertless retrovirus vector (−) independently in NB4-R1 (left panel) and NB4-S1 APL cells (right panel). Actin expression confirmed similar protein loading. (B) Retroviral-mediated knock-down of UBP43 triggered significant apoptosis in NB4-R1 (left panel) and NB4-S1 (right panel)APL cells. (C) Independent retroviral-mediated knock-down of UBP43 in NB4-R1 and NB4-S1 APL cells significantly inhibited proliferation as compared to an insertless retroviral vector. Significant changes are depicted (**, P < 0.005).

Fig. 5

Gain of UBP43 expression in APL cells. (A) Independent UBP43 immunoblot analyses were performed in the indicated APL cell lines engineered after UBP43 (+) or an insertless (−) retroviral transduction. Increased UBP43 expression augmented PML/RARα expression in NB4-R1 (left panel) and NB4-S1 APL cells (right panel). Actin expression confirmed similar protein loading. (B) UBP43 retroviral transduction inhibited apoptosis in NB4-R1 and NB4-S1 cells versus insertless control transduction, as measured by these caspase 3/7 assays. (C) Engineered gain of UBP43 expression augmented proliferation of NB4-R1 and NB4-S1 APL cells versus insertless vector controls. Significant changes are depicted (**, P < 0.005).

Fig. 6

RA-treatment of a murine transgenic transplantable APL model and of APL cells harvested directly from patients. (A) UBE1L, ISG15 and UBP43 mRNA expression profiles were each studied in RA-treated versus vehicle control-treated APL mice (see Materials and Methods). After RA-treatment for the indicated hours (h), UBE1L, ISG15 and UBP43 mRNAs were each increased in the isolated spleens. (B) RA-treatment induced UBE1L, ISG15 and UBP43 mRNA expression in cultured human APL cells from two representative RA-responsive cases, but not appreciably in a representative RA-resistant APL case. Expression profiles were studied after RA-treatment or no treatment and differentiation response was assessed by nitrotetrazolium blue (NBT) staining as described in the Materials and Methods. NBT positive cells appeared after RA-treatment of RA-responsive, but minimally in RA-resistant APL cells from these cases. Significant changes are indicated (**, P < 0.005).