Proteasome-dependent, ubiquitin-independent degradation of the Rb family of tumor suppressors by the human cytomegalovirus pp71 protein

Cell Culture.

The human osteosarcoma cell line, U-2 OS (American Type Culture Collection) was maintained in DMEM supplemented with 10% (vol/vol) FBS (HyClone) in an atmosphere of 5% CO₂ at 37°C. The mouse ts20 cell line, which contains a temperature-sensitive E1 ubiquitin-activating enzyme, and the H38 control cell line, in which the ts20 mutation was repaired, were gifts from H. Ozer (New Jersey Medical School, Newark, NJ). These cells were maintained at 35°C (the permissive temperature); their restrictive temperature was 39°C.

Degradation Analysis.

For cotransfection assays, U-2 OS cells were transfected with the indicated plasmids by the calcium phosphate method. The DNA precipitate was removed 12 h later, and cells were rinsed and cultured in 10% (vol/vol) FBS-containing medium for an additional 48 h before protein lysates were harvested at 60 h after transfection. The ts20 cells were transfected at 35°C on 10-cm plates with FuGENE 6 (Roche Applied Sciences) for 24 h, after which the cells were split into two dishes. After incubation at the permissive temperature (35°C) for an additional 12 h, some plates were moved to the restrictive temperature (39°C). Lysates were harvested 24 h later. Protein levels were quantified by the Bradford assay, polypeptides in lysates were separated by electrophoresis in an SDS-containing polyacrylamide gel, and Western blots were probed with the indicated antibodies. For the pulse–chase analysis, at 24 h after transfection, U-2 OS cells cultured in 10% (vol/vol) serum-containing medium were incubated in cysteine- and methionine-free medium containing 10% (vol/vol) dialyzed FBS for 1 h before a 30-min pulse with 4 ml of medium containing 0.2 mC/ml ³⁵S-Express (Perkin-Elmer). Cultures were chased in medium containing 10% (vol/vol) FBS, and lysates were immunoprecipitated with the hemagglutinin (HA)-specific antibody. Radioactive bands in blots were quantified by using a PhosphorImager (Molecular Dynamics).

Plasmids.

pSG5-pp71, pSG5-pp71-C219G, pSG5L-HApRb, pCMVp130, and pHAp130 have been described (refs. 23 and 24 and references therein). The pSG5 vector is from Stratagene. pHA-ubq (provided by H. Laman, Imperial Cancer Research Fund, London) encodes HA-tagged ubiquitin. pCW7 and pCW8 (provided by R. Kopito, Stanford University, Stanford, CA) encode yeast wild-type and dominant-negative (K48R mutant) ubiquitins, respectively (25).

Assay for in Vivo Ubiquitination.

Transfected U-2 OS cells were lysed in 100 μl of denaturation buffer (100 mM Tris, pH 7.4/1 mM EDTA/100 mM NaCl/2 mM DTT/3.5% SDS) with mild sonication and boiling (10 min). Debris was removed by centrifugation, and 5-μl aliquots were used for Western blot analysis. The remaining portions of denatured lysates were diluted to 6 ml in HK buffer (50 mM Hepes, pH 7.9/50 mM KCl/1 mM EDTA/0.5% Nonidet P-40/2 mM DTT/1 mM sodium orthovanadate/1 mM PMSF/25 mM NaF/10 μg/ml pepstatin A/10 μg/ml aprotinin/25 μg/ml leupeptin/25 μg/ml trypsin inhibitor) and incubated for 30 min on ice to allow for renaturation. Lysates were incubated with 20 μl of rabbit polyclonal antibody for 2 h. Immune complexes were captured on protein A agarose beads (Amersham Pharmacia), washed four times with lysis buffer and once with PBS. Bound proteins were separated by electrophoresis in an SDS-containing polyacrylamide gel and visualized by Western blot assay.

Antibodies and Inhibitors.

The following antibodies were obtained from commercial sources: tubulin (DM 1A, Sigma), HA (16B12, Babco, Richmond, CA), and p130 (C-20, Santa Cruz Biotechnology). The p53 antibody was 421, and the pp71 antibody was 4F7 (23). Lactacystin and E64 were from Calbiochem.

Degradation of the Rb Family of Proteins by pp71 Occurs Without Detectable Polyubiquitination.

The HCMV pp71 protein binds directly to Rb, p107 and p130 through an Rb-binding LxCxD motif, targeting them for degradation (23). When expression vectors for pp71 and an Rb family member are cotransfected into cells, the Rb family member accumulates to much lower levels than when it is cotransfected with either an empty vector, the adenovirus E1A protein, or HCMV pp65, a protein related in sequence to pp71, but unable to modulate the cell cycle (23). The reduced level to which the Rb family members accumulate in the presence of pp71 results from their increased turnover. We have shown that both Rb and p130 degradation by pp71 is inhibited by multiple proteasome inhibitors (23). Fig. 1 displays an experiment that confirms this earlier result within transfected U-2 OS cells; the pp71-induced degradation of the hypophosphorylated form of p130 was inhibited by the proteasome inhibitor lactacystin but not by E64, an inhibitor of nonproteasomal proteases. Thus, pp71-mediated degradation of p130 requires proteasome function.

The majority of proteasomal substrates are targeted for degradation by the addition of polyubiquitin chains (3, 4), a reaction that is catalyzed by a variety of classes of ubiquitin ligase complexes. To determine whether pp71 might act as a component of a ubiquitin ligase, we searched in vivo for pp71-induced polyubiquitinated species of the Rb family members.

The hyperphosphorylated form of p130 is thought to be degraded by the ubiquitin–proteasome pathway because its level drops dramatically as cells exit G₀ and progress through G₁, and it can be stabilized by proteasome inhibitors (26). Recently, p130 has been shown to be ubiquitinated in vitro when phosphorylated on serine 672, and mutation of this residue to alanine increased the protein's stability in vivo (27), providing additional evidence for its degradation in vivo by the ubiquitin–proteasome pathway. To probe the role of ubiquitination in the degradation of p130 in vivo, we transfected U-2 OS cells with expression plasmids encoding untagged p130 and HA epitope-tagged ubiquitin, immunoprecipitated p130 from denatured lysates with a polyclonal antibody to p130, and probed with an antibody specific for the HA epitope by Western blot assay to detect polyubiquitinated proteins. We detected polyubiquitinated species (Fig. 2A, lane 2) that migrate in the gels more slowly than p130 (as compared with nonubiquitinated, HA-tagged p130, Fig. 2A, lane 1), providing direct evidence that p130 is ubiquitinated in vivo before degradation by the proteasome. Because pp71 increased the rate of p130 turnover, we expected that, if pp71 acted as a ubiquitin ligase, we would detect an increase in polyubiquitinated p130 in the presence of pp71. However, not only did we not detect an increase, we were unable to detect any polyubiquitinated species of p130 in the presence of pp71 (Fig. 2A, lane 3), in experiments where pp71 was shown to decrease the levels of p130 but not totally eliminate it from the lysate (Fig. 2C). This does not seem to be the result of a global inhibition of ubiquitination by pp71 because roughly similar levels of total ubiquitinated proteins were observed both in the presence and absence of pp71 (Fig. 2B).

Inhibition of proteasome function can stabilize some polyubiquitinated species and cause them to accumulate, whereas others become deubiquitinated upon prolonged proteasome inhibition. To determine whether proteasome inhibition could stabilize a putative pp71-induced polyubiquitinated p130 intermediate, we looked for such a species in lysates of transfected cells that were treated with lactacystin for either a short or long period. In the absence of pp71, we observed a small increase in polyubiquitinated p130 after 5 h (Fig. 3A) and a substantial increase after 14 h of lactacystin treatment (Fig. 3B). However, we did not detect polyubiquitinated species of p130 in the presence of pp71 in cells treated with lactacystin for either time period (Fig. 3), and, as before, roughly similar levels of total ubiquitinated proteins were observed in the presence and absence of pp71 (Fig. 3A). Thus, we cannot detect polyubiquitination of p130 in the presence of pp71.

Conceivably, pp71 degrades hypophosphorylated p130 in the absence of polyubiquitination and before it can become phosphorylated when it would be subject to normal cell cycle-mediated polyubiquitination. In support of this hypothesis, we can detect polyubiquitinated p130 in the presence of a pp71 variant, C219G (Fig. 3B), whose mutation resides in the Rb-binding LxCxD motif. This mutant does not stimulate cell cycle progression out of G₀, and it fails to degrade Rb and p130 (23). Polyubiquitinated species of p130 are observed in the presence of the C219G pp71 protein, presumably because degradation of the hypophosphorylated form of p130 is prevented. This could allow p130 to be phosphorylated and hence become a substrate for a cellular ubiquitin ligase. We also were unable to detect polyubiquitinated species of Rb or p107 either in the presence or absence of pp71, with or without proteasome inhibition (data not shown). Because Rb and p107 are thought to be regulated only by phosphorylation and not by targeted destruction, and because pp71 does not seem to function as a ubiquitin ligase, this was the expected result.

Rb Family Degradation by pp71 Is Maintained During Conditions That Inhibit the Polyubiquitination of Proteins.

Our analysis of the ubiquitination status of p130 points to a proteasome-dependent, ubiquitin-independent mechanism for the pp71-mediated degradation of the Rb family. An alternative explanation is that pp71-induced ubiquitination of p130 may be undetectable even in this sensitive assay because of technical reasons, such as the low levels of p130 present in the lysates, the short-lived nature of a putative pp71-induced polyubiquitinated p130 intermediate, or the fact that such a species is not stabilized by lactacystin. To distinguish between these two models, we used two independent methods to determine whether polyubiquitination is required for the pp71-mediated degradation of the Rb family.

In the first approach, we asked whether a dominant-negative ubiquitin molecule (DN-ubq) could inhibit the pp71-mediated degradation of p130. Ubiquitin chains are extended by the covalent polymerization of ubiquitin molecules through their terminal glycine residue to the lysine at amino acid position 48 of the last ubiquitin in the existing chain. When added to existing chains, a ubiquitin mutant with lysine 48 replaced by arginine, K48R, exerts a dominant-negative, chain-terminating effect and has been shown to inhibit the proteasomal degradation of several proteins (9, 25, 28). We chose to examine p130 because, in the absence of pp71, it is polyubiquitinated and degraded by the proteasome and thus provides an internal positive control in this assay. A pulse–chase protocol was used because competition between exogenous DN-ubq and the endogenous wild-type molecule for incorporation into chains often leads to substantially less than 100% inhibition (25) and thus might not be evident by Western blot analysis. In the absence of pp71, an inhibitory effect on p130 degradation of DN-ubq was observed when compared with cotransfection with an expression plasmid for wild-type ubiquitin (Fig. 4A Upper). However, DN-ubq had no effect on the pp71-mediated degradation of p130 (Fig. 4A Lower), which, as previously illustrated (23), occurred at a much faster rate than in the absence of pp71 (data not shown). In three independent experiments, we observed an ≈50% inhibition of p130 degradation by DN-ubq in the absence of pp71 and less than 2% inhibition in the presence of pp71 (Fig. 4B). As DN-ubq cannot inhibit the pp71-mediated degradation of p130, the simplest conclusion is that this degradation occurs in the absence of polyubiquitination.

Finally, we performed degradation experiments in the mouse ts20 cell line (29). These cells have a temperature-sensitive E1 ubiquitin-activating enzyme, which is nonfunctional at the restrictive temperature of 39°C. At this temperature, ubiquitin monomers cannot be activated, polyubiquitin chains are not synthesized, and proteins degraded by the ubiquitin–proteasome pathway, such as p53, are stabilized (29). In these cells, we were able to detect the pp71-mediated degradation of the hypophosphorylated form of Rb at the restrictive temperature, under conditions where the p53 protein was stabilized (Fig. 5). Because, under these conditions, polyubiquitination is prevented but pp71-mediated degradation of Rb still occurs, we conclude that this degradation occurs through a ubiquitin-independent mechanism.