Arsenite-induced Cdc25C degradation is through the KEN-box and ubiquitin–proteasome pathway

Cell Culture.

The human bronchial epithelial cell line, BEAS-2B, from American Type Culture Collection, was cultured as previously described (10). To determine the rate of cell proliferation, cells were seeded in six-well tissue culture plates at ≈3 × 10⁴ cells per well with or without arsenite. At each time point, wells in triplicate were rinsed with PBS twice to remove dead cells and debris. Live cells on the plates were trypsinized and counted by using a hemocytometer. For the trypan blue dye exclusion assay, all cells were collected separately from each well at the time indicated. The cells were stained with trypan blue (Sigma) for 10 min. The stained (dead) and unstained (live) cells were counted. For flow cytometry, cells were labeled with propidium iodide and subjected to cell cycle profile analysis, as described previously (10).

Reagents.

Arsenite was purchased from Aldrich. Antibodies against human Cdc25A, Cdc25B, Cdc25C, p55Cdc20, and ubiquitin were from Santa Cruz Biotechnology or Upstate Biotechnology (Lake Placid, NY). Enhanced chemiluminescence Western blotting detection reagents were from Amersham Pharmacia Life Science.

Tat-Cdc25C Peptides.

A modified 11-mer Tat transduction domain (single-letter code, YGRK/AK/ARRQRRR) was used as a delivery sequence for wild-type Cdc25C peptide 141–160 encompassing a KEN box or a mutated Cdc25C peptide in which the KEN box was replaced with an AAA sequence. Because of a concern that the presence of lysine residues in the delivery peptide derived from HIV Tat protein would interfere with ubiquitination of intracellular proteins, two lysine residues (underlined in the indicated Tat-transduction domain) were replaced with alanines in the delivery peptide. The Tat-Cdc25C-KEN peptide contains a sequence of YGRAARRQRRR-G-RDAMCSSSANKENDNGNLVD. The Tat-Cdc25C-AAA peptide contains a sequence of YGRAARRQRRR-G-RDAMCSSANAAADNGNLVD. FITC-labeled Tat peptide was generated by fluorescein labeling. All peptides were purified by gel filtration through fast protein liquid chromatography. The purity of peptides was around 95%.

Immunoprecipitation and Western Blot.

Cells were trypsinized, washed twice with ice-cold PBS, and harvested in cell lysis buffer containing the protease inhibitors antipain (10 μg/ml), pepstatin A (10 μg/ml), and leupeptin (10 μg/ml). Cells were lysed by sonication, and the protein supernatant was clarified by centrifugation at 25,000 × g for 5 min at 4°C. Protein concentration was determined by the Bradford protein assay. Immunoprecipitation was performed by adding anti-Cdc25C, antiubiquitin, or anti-Myc-tag antibody. Western blot analysis was performed by using the following primary antibodies: anti-checkpoint kinase 1 (Chk1) polyclonal antibody, anti-Cdc25A polyclonal antibody, anti-Cdc25B polyclonal antibody, anti-Cdc25C polyclonal antibody, anti-p55Cdc20, and antiubiquitin polyclonal antibody.

Site-Directed Mutation.

Point mutation was introduced into plasmid pCMV-Myc-Cdc25C by using the QuickChange Site-Directed Mutagenesis kit (Stratagene). To generate a K to A mutation for the first lysine (K) residue in the KEN box of Cdc25C, the coding sequence AAA for lysine was changed to GCA by using a primer set encompassing the coding sequence of KEN box.

Arsenite Treatment Causes Proliferative Arrest in BEAS-2B Cells.

As described earlier (12, 13), arsenite is a potent inducer of stress responses of the cell, which possibly affect cell proliferation or differentiation. To determine the further effect of arsenite on cell growth, the cells were visually inspected daily after 5 or 10 μM arsenite treatment. Control cells showed no decrease in cell density over a 5-day period. In contrast, cells treated with arsenite showed a progressive decrease of cell density because of the detachment of dead cells. Treatment of cells with arsenite suppressed the proliferation of BEAS-2B cells that had been proliferating asynchronously, as indicated by cell numbers (Fig. 1A), whereas mock treatment had no effect (Fig. 1A). The reduced growth of the cells treated with arsenite appeared to be due to an increased rate of cell death. Trypan blue dye exclusion analysis indicated a pronounced decrease in cell viability beginning at day 2, which became prominent by day 5 (Fig. 1B). Treatment of the cells with arsenite for 2 days promoted the detachment of the cells. More than 90% of detached cells were stained by trypan blue.

Arsenite Causes Cell-Cycle Arrest in G₂/M Phase.

To explore further the inhibitory effect of arsenite in human bronchial epithelial cell proliferation, we analyzed the cell cycle profile in the absence or presence of arsenite for 48 h by measuring the DNA content by flow cytometry analysis of propidium-iodide-stained cells (Fig. 2A). Because previous studies showed that arsenite was able to increase the number of asynchronized cells in both S phase and G₂/M phase, we speculated that the arsenite effect might be more pronounced if cells were synchronized in G₀ phase through prestarvation in serum-free medium for 36 h. The cells were cultured in the absence or presence of arsenite in serum-free medium for an additional 48 h after prestarvation. Note that serum starvation for an additional 48 h resulted in more than 60% of cells accumulating in G₀/G₁ phase. In contrast, over 50% of the cells were in G₂/M phase after 48 h of sustained treatment of the cells with 5–10 μM arsenite. An average cell cycle distribution of three independent experiments was presented in Fig. 2B, which indicated that the majority of the cells were in G₀/G₁ in the absence of arsenite. Arsenite treatment resulted in a dose-dependent decrease of G₀/G₁ cells concomitant with an increase of G₂/M cells. In addition, a small (statistically insignificant) increase of S phase cells in response to arsenite was observed. This result suggests that the S phase entry of the cells was normal in response to arsenite, but the exit from S phase might be partially impaired.

Arsenite-Induced G₂/M Arrest Involves Cdc25C Degradation.

We presumed that arsenite-induced G₂/M arrest could be attributed to the activation of a checkpoint control. Gross overproduction of checkpoint proteins is rarely seen in normal cells because of the powerful growth-suppressing capabilities of these proteins. However, increased checkpoint protein expression or activity can often be observed after the treatment of the cells with DNA-damaging agents or other stress inducers. A number of checkpoint proteins, including GADD45α, p53, Chk1, and Chk2, can play a role in regulating the G₂/M phase checkpoint. Because the function of p53 is deficient in BEAS-2B cells as previously reported (14), other checkpoint proteins that could account for arsenite-induced G₂/M phase arrest were investigated.

Our previous work demonstrated that arsenite was able to induce GADD45α expression, especially in cells where NF-κB signaling was inhibited (10, 11). In cells where NF-κB signaling was normal, the induction of GADD45α by arsenite was less potent, whereas the G₂/M arresting effect of arsenite was still substantial, indicating the involvement of other checkpoint protein(s). We assumed that Chk1 and Chk2 might account for this G₂/M arrest in addition to GADD45α. In response to DNA damage or other stress signals, activated Chk1 and Chk2 phosphorylate S216 of Cdc25C, causing it to bind to 14–3-3 proteins, which retain Cdc25C in the cytoplasm where it cannot dephosphorylate and activate Cdc2. As a consequence, cells will be arrested in G₂/M phase. Functional induction of Chk1 and Chk2 can be assessed by examining the protein levels and mobility of Cdc25C. When cells were cultured in the absence of arsenite, a steady level of Cdc25C was observed both in the continually serum-starved condition and in the presence of serum (5%) (data not shown). A rapid degradation of Cdc25C was observed after incubation of cells with 5 μM arsenite for different time periods (Fig. 3A). The earliest degradation of Cdc25C occurs at 30 min of arsenite stimulation. The arsenite-induced degradation of Cdc25C occurs in a dose-dependent manner (Fig. 3B Top). Because nonspecific protein degradation occurs when cells undergo cytotoxic responses, we determined whether the degradation of Cdc25C induced by arsenite represented a cytotoxic reaction. Thus, the protein levels of two other Cdc25 family proteins, Cdc25A and Cdc25B, were examined. Whereas a marginal degradation pattern of Cdc25B was observed in the cells treated with arsenite (Fig. 3B Middle), a dose-dependent increase of Cdc25A was noted (Fig. 3B Bottom). In contrast to the regulation of G₂/M by Cdc25B and Cdc25C, the major function of Cdc25A is the regulation of the G₁/S phase transition. The increased Cdc25A protein not only explains the observed small increase in the number of cells in S phase (Fig. 2 A and B) but also excludes the possibility that nonspecific cytotoxicity caused the observed degradation of Cdc25C in response to arsenite. Indeed, cytotoxicity did not occur until 24 h after the cells were treated with arsenite (Fig. 1).

Arsenite-Induced Cdc25C Degradation Depends on Ubiquitin–Proteasome.

Emerging evidence indicates that ubiquitin–proteasome participates in the degradation of several cell cycle regulatory proteins, such as cyclin A, cyclin B, Cdc25A, Cdc25B, Cdc6, geminin, and securin/pituitary tumor-transforming gene (PTTG) (6). To test whether the degradation of Cdc25C induced by arsenite is mediated through ubiquitin–proteasome or other systems, cells were preincubated with a specific proteasome inhibitor, lactacystin, or a calpain inhibitor, calpastatin. Analysis of Cdc25C by immunoblotting of lysates from cells pretreated with lactacystin revealed that the degradation of Cdc25C protein induced by arsenite was inhibited in the presence of lactacystin (Fig. 4A; compare lane 2 with lane 3). In contrast, calpastatin peptide, a specific cell-permeable inhibitor for calpains, had no effect on arsenite-induced Cdc25C degradation (Fig. 4A, lane 4). Fig. 4A also shows that multiple conjugates with higher molecular weights than Cdc25C itself can be recognized by the anti-Cdc25C antibody. To test whether some of these conjugates could form from polyubiquitin modification on Cdc25C, we monitored the possible presence of ubiquitin conjugation by reprobing the stripped membrane with antiubiquitin antibody. Results depicted in Fig. 4B indicate that high molecular-weight conjugates can be recognized by the antibody against ubiquitin. The ubiquitination of Cdc25C induced by arsenite appears to be very rapid, because a time-course study indicates that the ubiquitination of Cdc25C occurred after as little as 30 min of arsenite stimulation in the presence of lactacystin (Fig. 4C). Thus, these results suggest that arsenite-induced degradation of Cdc25C is through the ubiquitin–proteasome system.

To confirm further the ubiquitination of Cdc25C in response to arsenite, we immunoprecipitated Cdc25C or ubiquitin from the lysates of cells treated with arsenite in the presence of lactacystin and analyzed the samples by immunoblotting. Ubiquitin conjugation with Cdc25C was confirmed by a reciprocal experiment in which proteins were first immunoprecipitated with Cdc25C antibody and analyzed by Western blot with the ubiquitin antibody (Fig. 4D, lanes 1 and 2) or vice versa (Fig. 4D, lanes 3 and 4).

Arsenite-Induced Cdc25C Degradation Requires KEN Box.

The ubiquitin–proteasome-mediated degradation of many cell cycle regulatory proteins requires either a D box composed of the sequence R-X-X-L-X-X-X-X-N or a KEN box with amino acid sequence K-E-N-X-X-X-N/D (9, 15, 16). Analysis of the amino acid composition of the human Cdc25C revealed the absence of a typical D box in its sequence. However, there is a KEN-like structure with a sequence K-E-N-D-N-G-N in the region of 151 to 157 of Cdc25C. Sequence alignment among several vertebrate Cdc25Cs shows that all Cdc25Cs analyzed contain a roughly conserved KEN box (Fig. 5A). To determine whether the KEN box within the region of 151 to 157 is required for the ubiquitination of Cdc25C, we synthesized a Tat-fusion peptide containing an 11-mer Tat transduction domain fused to a 20-aa peptide encompassing the KEN box of human Cdc25C or the control sequence in which the KEN residues were replaced with AAA. We used the Tat-Cdc25C-KEN box peptide as a competitor to inhibit the ubiquitination of Cdc25C protein in response to arsenite. To monitor the efficiency of Tat fusion peptides to transduce into cells, a Tat peptide conjugated to FITC was added to the cell culture medium in a concentration of 20 μM for 2 h. Consistent with previous reports indicating the efficient cellular entry of Tat-fusion peptides (17), fluorescein microscopic analysis demonstrated that more than 95% of cells were transduced with Tat-FITC peptide (Fig. 5B). Microscopy at higher magnification indicates that most of Tat-FITC peptides were accumulated in the nuclei (data not shown). To determine the effect of Tat-Cdc25C-KEN box on ubiquitination, cells were pretreated with lactacystin and 20 μM Tat-Cdc25C-KEN box or control peptide. After 2 h, cells were further treated with increasing concentrations of arsenite for 12 h. Cell lysates were immunoprecipitated and analyzed for ubiquitination of Cdc25C. The results showed that the Tat-Cdc25C-KEN-box peptide but not the control peptide, KEN-box-mutated Tat-Cdc25C-AAA peptide, was capable of inhibiting the ubiquitination of Cdc25C in the cellular response to increasing concentrations of arsenite (Fig. 5C; compare lanes 6–10 with lanes 1–5). Alternatively, we pretreated cells with lactacystin along with increasing concentrations of Tat-KEN-box peptide or control peptide for 2 h. A decreased arsenite-induced Cdc25C ubiquitination was observed in the cells in response to the increasing concentrations of Tat-KEN-box peptide (Fig. 5D, lanes 1–5) but not the increasing concentrations of control peptide (Fig. 5D, lanes 6–10).

To ascertain directly the requirement of KEN box for the ubiquitination of Cdc25C induced by arsenite, we generated Myc-tagged Cdc25C expression vector with a mutation of KEN box (K151A). Transfection of this mutated form of Cdc25C revealed that the mutant exhibited the same mobility and expression level as the wild-type form of Cdc25C (Fig. 5E). However, the ubiquitination of this mutated Cdc25C in response to arsenite was reduced appreciably in the immunoprecipitation by using anti-Myc-tag antibody. The remaining ubiquitin conjugation on the K151A mutated form of Cdc25C indicates that other regions in Cdc25C protein might also contribute, although less effectively, to the arsenite-induced ubiquitination of Cdc25C.

Arsenite Induces Degradation of Cdc20.

The requirement of KEN box for arsenite-induced Cdc25C degradation suggests that arsenite may target other KEN-box-containing proteins for proteasomal degradation. The degradation of Cdc20, an activator of mitotic form APC, had been previously shown to be KEN box-dependent (9). To determine whether arsenite is also capable of inducing the degradation of Cdc20, we next assayed the ubiquitination and protein levels of Cdc20 in cells treated with arsenite. As depicted in Fig. 6, arsenite induced degradation of Cdc20 in both a dose- and time-dependent manner. However, in comparison with Cdc25C (Figs. 3 and 4), both ubiquitination and degradation of Cdc20 are relatively weak in response to arsenite.

S216 Phosphorylation of Cdc25C Is Not Involved in Arsenite-Induced Cdc25C Degradation.

It has been well documented that stress signals activate a subset of cellular kinases, such as mitogen-activated protein kinase p38 and Chk1, that phosphorylate S216 of Cdc25C (4, 18). Modification of Cdc25C by S216 phosphorylation regulates the binding of 14–3-3 proteins that sequester Cdc25C in the cytoplasm. It has been shown that phosphorylation of Cdc25A by Chk1 is an essential step for the ubiquitin–proteasome-mediated degradation of Cdc25A in the cellular response to DNA-damaging agents (19). To determine whether phosphorylation of S216 affects the ubiquitination and degradation of Cdc25C induced by arsenite, we transfected a Myc-tagged Cdc25C with S216A mutation. As indicated in Fig. 7 A and B, the mutation of S216 had no appreciable effect on the ubiquitination or degradation of Cdc25C protein in the cellular response to arsenite. Our previous studies showed a substantial activation of mitogen-activated protein kinases (MAPK), i.e., Erk, c-Jun N-terminal kinase (JNK), and p38, by arsenite (10, 11). A recent study by Bulavin et al. (18) suggested that genotoxic activation of p38 was responsible for the S216 phosphorylation of Cdc25C and subsequent G₂/M cell cycle arrest. To determine whether activation of MAPK contributes to arsenite-induced Cdc25C degradation, cells were preincubated with an Erk kinase inhibitor (PD98059), a p38 inhibitor (SB203580), or transfected with a kinase mutated SEK1 to block JNK activation. Analysis of the protein levels of Cdc25C in the cells treated with arsenite indicates that none of these inhibitors could restore the steady level of Cdc25C (data not shown), despite their effective inhibition on arsenite-induced activation of Erk, JNK, and p38, as previously described (10, 11).