Von Hippel-Lindau–Coupled and Transcription-Coupled Nucleotide Excision Repair–Dependent Degradation of RNA Polymerase II in Response to Trabectedin

Natural products are a rich source for medicinal drugs endowed with remarkable ability to selectively target biomolecules (1). Ecteinascidin 743 (Et743; trabectedin) was originally purified from the marine tunicate Ecteinascidia turbinata in 1990 (2) and developed because of its activities in experimental cancer models (2–6). Et743 has recently been approved in the European Union for the treatment of soft tissue sarcomas and is in clinical trials for ovarian, breast, and prostate cancers and for pediatric sarcomas. Et743 has been granted Orphan Drug designation from the European Commission and the U.S. Food and Drug Administration for soft tissue sarcomas and ovarian cancer.³

The mechanism of action of Et743 is unique in that the antiproliferative activity of Et743 is dependent on transcription-coupled nucleotide excision repair (TC-NER; see refs. 7–9 for reviews). Et743 is also a potent transcription inhibitor (10–14). It binds in the DNA minor groove and alkylates the exocyclic N2 position of guanines with a preference for guanines that are 5′ from another guanine or a cytosine (15, 16). As it widens the minor groove, Et743 bends the DNA sharply toward the major groove opposite to its alkylation site (15, 16). Et743-DNA adducts then arrest RNA polymerase II (Pol II), which recruits TC-NER complexes. However, instead of excising the Et743 adduct, TC-NER complexes become trapped while attempting to process the Et743-mediated DNA damage (17–19).

NER is subdivided in two pathways depending whether the DNA adducts are repaired in transcribing (TC-NER) or nontranscribing DNA [global genome NER (GG-NER); see refs. 20, 21 for review]. In the case of TC-NER, the adducts block the progression of Pol II, and the stalled Pol II complexes act as recognition complexes for NER and activate the TC-NER–specific factors Cockayne syndrome groups A and B (CSA and CSB). CSA and CSB, respectively, correspond to the two complementation groups for cells derived from patients with inherited CS. CSB is the ortholog of the yeast Rad26 gene (22). In the TC-NER pathways, CSB/Rad26 recruits the XP repair complex, thereby initiating NER. It also induces the ubiquitylation-degradation of Pol II, which provides access for the NER complex (23). In the second NER pathway (GG-NER), adducts in nontranscribing DNA are recognized by the XPC-hHRN23B heterodimer, which recruits the other XP factors (20, 21, 24, 25). The following steps are common for both TC-NER and GG-NER. XPA first binds to the DNA damage recognition complex. The XPD and XPB DNA helicases, which are components of the transcription factor TFIIH complex, then unwind the two DNA strands at the damaged site. XPG and XPF are nucleases that incise the adducted strand in the late stage of NER (20, 21). The Et743 adducts have been proposed to trap dead-end XPG-DNA complexes (17, 26), thereby generating DNA single-strand breaks (17) selectively in the TC-NER subpathway. The selectivity of Et743 for TC-NER versus GG-NER might be due to the selective recognition of the Et743-DNA adducts by TC-NER. It is also not excluded that the complexes resulting from the association of XPG with TFIIH and a stalled RNA Pol II complex might be preferentially trapped by Et743 because of differential overall structure of such complexes compared with the GGR complexes.

Transcription arrest during Pol II–mediated RNA elongation not only is a sensor for TC-NER but also poses functional problems for DNA repair. Pol II consists of a large multiprotein complex, which needs to be removed for the TC-NER complex to gain access to the DNA damaged site. This can be accomplished by at least two mechanisms: (a) backward movement (“back tracking”) of Pol II and (b) ubiquitylation-degradation of Pol II (23). It is not well understood when these mechanisms operate in response to transcription-blocking lesions. The molecular identity of the ubiquitin-proteasome pathways involved in Pol II degradation also remained to be fully determined. A recent study in yeast showed that polyubiquitylation and degradation of the large subunit of Pol II (Rpb1) is mediated by the ubiquitin ligase (E3), which is composed of elongins C and A, cullin3, and the RING finger protein Roc1 (Rbx1), and suggested that the von Hippel-Lindau (VHL) tumor suppressor complex might promote ubiquitylation and degradation of Rpb1 in human cells (22).

The present study was undertaken to elucidate the molecular effect of trapping abortive TC-NER by Et743 in human cancer cells. We find that Et743 is remarkably effective in inducing Pol II degradation in a variety of cell lines. We also find that Pol II degradation is TC-NER and VHL dependent and that it is coordinated with cell killing by Et743.

Drugs and antibodies. Et743 was kindly provided by PharmaMar. Z-Leu-Leu-Leu-al (MG132) and 5,6-dichlorobenzimidazole-1-β-d-ribofuranoside (DRB) were purchased from Sigma. The following antibodies were used for Western blotting: Pol II (clone N20), phospho-Ser⁵ Pol II (clone 8A7), XPD (TFIIH p80, clone H150), and CSB (clone E18) from Santa Cruz Biotechnology and phospho-Ser² Pol II (ab5095) from Abcam. Phospho-Ser² Pol II antibody (clone H5) from Covance was used for immunofluorescence microscopy.

Cell lines. The human colon carcinoma HCT116 and HT29 cell lines and prostate carcinoma cell line DU145 were obtained from the Developmental Therapeutics Program (National Cancer Institute, NIH, Bethesda, MD). The human fibroblast GM00637, XPF, and XPG cell lines were obtained from Coriell Cell Repository. Human CSB fibroblasts (CS1AN.S3.G2) complemented with CSB (CSB-C) or transfected with empty vector (CSB-V) have been described previously (27). Human fibroblast XPD and XPC cell line and their complemented XPD-C and XPC-C sublines were provided by Dr. Kenneth Kraemer (Basic Research Laboratory, NIH, Bethesda, MD). Ewing's sarcoma TC-32 cells were provided by Dr. Crystal Mackall (Pediatric Oncology Branch, NIH, Bethesda, MD). The renal cell carcinoma cell line 786-0 contains a single VHL allele with a frameshift mutation at codon 104 (28). Stable transfectants of 786-0 (786-0-wt) and vector-containing truncate VHL construct have been described previously (28). VHL mutants and stable transfectants were provided by Dr. Marston W. Linehan (Urologic Oncology Branch, Center for Cancer Research, NCI, NIH, Bethesda, MD). All the above cell lines were cultured in DMEM supplemented with 10% fetal bovine serum (Gemini Bio-Products). CSB-C and CSB-V cells were additionally supplemented with 400 μg/mL geneticin (Invitrogen). All experiments were done in exponentially growing cells with a density not exceeding 30%. In confluent cells, Pol II degradation was reduced and Pol II levels remained detectable over several hours (data not shown). Thus, the present study, which focuses on the mechanisms of Et743-induced Pol II down-regulation, was done in exponentially growing cells.

Western blotting. Whole-cell extracts were prepared by lysing the cells at 4°C in 1% SDS, 1 mmol/L sodium vanadate, and 10 mmol/L Tris-HCl (pH 7.4) in the presence of protease inhibitors (Complete, Roche Diagnostics) and phosphatase inhibitors (phosphatase inhibitor cocktail 1, Sigma). Viscosity of the samples was reduced by brief sonication. Proteins (50 μg; Bio-Rad detergent-compatible protein assay) were mixed in loading buffer [125 mmol/L Tris-HCl (pH 6.8), 10% β-mercaptoethanol, 4.6% SDS, 20% glycerol, 0.003% bromphenol blue], separated by SDS-polyacrylamide gel, and transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore). After blocking nonspecific binding with 5% milk in TPBS (0.2% Tween 20, PBS) for 1 h, membranes were incubated for 2 h with primary antibody. After further washes in TPBS, membranes were incubated with horseradish peroxidase–conjugated goat anti-rabbit or anti-mouse antibody (1:10,000 dilution; Amersham Biosciences) for 1 h and then washed in TPBS. Immunoblots were revealed using enhanced chemiluminescence detection kit (Pierce) followed by autoradiography.

Confocal microscopy analyses of Pol II phosphorylated at Ser². HCT116 cells were grown in Nunc chamber slides (Nalgene). Following treatment, cells were washed in PBS, fixed with 2% paraformaldehyde, postfixed, and permeabilized with 70% ethanol. After blocking nonspecific binding sites with 8% bovine serum albumin for 1 h, cells were incubated for 2 h with anti-phospho-Ser² Pol II antibody (clone H5) at 1:100 dilution and tagged for 1 h with fluorescent secondary antibodies (Molecular Probes). Slides were mounted using Vectashield mounting liquid containing propidium iodide (Vector Laboratories) and visualized with a Nikon Eclipse TE-300 confocal laser scanning microscope system.

Cell survival assays. Exponentially growing renal carcinoma cells were seeded in 96-well plates on day 1 (1,500 per well) and allowed to attach overnight. Et743 was added on day 2 and cell cultures were exposed continuously for 72 h. After exposure to Et743, 50 μL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reagent (Sigma) were added directly to the 96-well plates and incubated for 4 h. After removal of culture medium by aspiration, 100 μL DMSO was added to each well to dissolve the MTT. The amount of MTT colorimetric product produced in each well was assessed by reading the absorption of UV light at 650 nm on an Emax Precision Microplate UV spectrophotometer (Molecular Devices). For determination of growth inhibition, the six data points for each drug concentration were averaged and expressed as percentage of untreated control. Each experiment was done a minimum of three times on separate days and all resulting data were imported into GraphPad Prism 4.0 (GraphPad Software) for statistical analysis and visual representation. Plotted data represent the mean result of the individual experiments with accompanying SE.

Et743 induces Pol II down-regulation. Transcribing Pol II is characterized by hyperphosphorylation of the COOH-terminal domain of Rpb1 (its largest subunit) on Ser², Ser⁵, and Ser⁷ of canonical heptapeptide repeats Tyr¹-Ser²-Pro³-Thr⁴-Ser⁵-Pro⁶-Ser⁷ (29–31) and can be distinguished from hypophosphorylated Pol II because of its reduced electrophoretic mobility in SDS polyacrylamide gels (32). Thus, Pol IIo corresponds to the actively transcribing Pol II, whereas Pol IIa corresponds to the nontranscribing Pol II (29). Figure 1 shows that under normal conditions, both Pol IIa and Pol IIo can be detected in human colon carcinoma HCT116 cells (Fig. 1A, left lane). Figure 1 also shows that treatment with nanomolar concentrations of Et743 induces the disappearance of both Pol IIa and Pol IIo (Fig. 1A and B). This decrease was rapid and massive as most Pol II disappeared within 30 min in cells exposed to 10 nmol/L Et743 (Fig. 1B). Et743-induced Pol II disappearance was observed in various human cell types, including Ewing's sarcoma, renal cancer cells, and transformed fibroblasts (see Figs. 2– 4 and Table 1).

To determine whether the disappearance of Pol II in response to Et743 was related to active transcription, we tested whether the transcription inhibitor DRB could prevent Et743-induced Pol II loss. DRB prevents the engagement of Pol II in transcription elongation by inhibiting the cyclin-dependent kinases Cdk9 and Cdk7, which are essential for phosphorylating Pol II COOH-terminal repeats (33, 34). As expected, DRB alone promoted the decrease of the hyperphosphorylated actively transcribing Pol II (Pol IIo; Fig. 1C, lane 2; ref. 29). Figure 1C also shows that Et743 failed to induce the disappearance of Pol II in cells treated with DRB. Thus, these experiments indicate that Et743 promotes transcription-dependent down-regulation of Pol II.

Et743 induces proteasome-dependent degradation of transcribing Pol II. Because Et743 is remarkably active in soft tissue sarcoma and Ewing's sarcoma cells (35–37) and has recently been approved for the treatment of soft tissue sarcomas, we tested the effects of Et743 on Pol II down-regulation in human sarcoma cells. Figure 2A shows rapid and massive Pol II down-regulation in Ewing's sarcoma TC-32 cells treated with 10 nmol/L Et743.

In those cells, we also examined whether Pol II down-regulation was reversible. Figure 2B shows that following a 1-h treatment with 10 nmol/L Et743, Pol II levels remained low for several hours and only became detectable 8 h after Et743 removal. Thus, the Pol II down-regulation induced by Et743 is relatively persistent.

Next, we asked whether the loss of Pol II was related to its proteasomal degradation. Figure 2C shows that the proteasome inhibitor MG132 prevented Pol II down-regulation, indicating that Et743-induced Pol II down-regulation is due to its proteasomal degradation. Furthermore, Fig. 2C shows that Et743 promotes Pol II hyperphosphorylation in MG132-treated cells, as shown by the decrease of Pol IIa with coincident increase of Pol IIo (compare lanes 5 and 6 with lane 1). Experiments done using phosphospecific antibodies further showed that Pol II hyperphosphorylation was preferential at Ser⁵ in the heptapeptide repeats of Pol II, whereas Et743 produced no significant change in Ser² phosphorylation (Fig. 2C). These observations suggest that there is preferential proteasomal degradation of the transcriptionally engaged Pol II (Pol IIo; ref. 31) in cells treated with Et743.

Pol II degradation requires functional TC-NER. Because the antiproliferative activity of Et743 is dependent on TC-NER (17), we compared Pol II degradation in NER-deficient and NER-proficient cells. Figure 3A shows the effects of Et743 on Pol II in the XPD and XPD-C isogenic cell lines (middle and right). Pol IIo remained stably expressed and even increased in the XPD cells, whereas Pol II was rapidly degraded in the XPD-C cells. Hypophosphorylated Pol II (Pol IIa) decreased in the XPD cells but that decrease was associated with hyperphosphorylation of Pol II (increased Pol IIo; Fig. 3A, middle). That effect was similar to that of the proteasome inhibitor MG132 (see Fig. 2C). Thus, XPD cells fail to degrade Pol II but remain able to induce hyperphosphorylation of Pol II in response to Et743. To further establish the involvement of the NER in Pol II degradation, we examined additional NER-deficient cell lines. Et743-induced Pol II degradation was also defective in NER-deficient XPA, XPG, and XPF cells (Table 1).

Because of the selectivity of Et743 for the transcription-coupled subpathway of NER (TC-NER) over the GG-NER subpathway (17), we tested Pol II degradation in cells deficient in each of those two subpathways. CS results from a selective defect in the TC-NER pathway (38). Figure 3B compares Et743-induced Pol II degradation in a CSB-deficient (CSB-V) and complemented cells (CSB-C; left). CSB-deficient (CSB-V) cells showed defective Pol II degradation in response to Et743, whereas the complemented CSB-C rapidly degraded Pol II (Fig. 3B, middle and right). Moreover, the absence of Pol II degradation in CSB-deficient cells was associated with accumulation of hyperphosphorylated Pol II (Pol IIo; Fig. 3B, middle), which is consistent with the results obtained in XPD cells (see Fig. 3A). This result indicates that TC-NER–deficient CS cells behave like the XPD, XPA, XPF, and XPG cells, which are deficient for the common NER pathway (Table 1). The TC-NER–specific deficiency in Pol II degradation was also examined by immunofluorescence microcopy using a phosphospecific antibody against Pol II (Fig. 3C). In agreement with the Western blotting results shown in Fig. 3B, Pol II remained readily detectable in CSB-V cells treated with Et743, whereas the Pol II signal was strongly reduced in CSB-C treated with Et743. By contrast, the GG-NER–deficient XPC cells and their complemented counterpart (XPC-C; ref. 17) both degraded Pol II in response to Et743 (Fig. 3D). Together, these experiments show that TC-NER is required for Pol II degradation in response to Et743.

Pol II degradation involves VHL and leads to cell death in Et743-treated cells. Because recent studies suggested that the VHL tumor suppressor protein–associated complex could bind hyperphosphorylated Rpb1 and the Rpb7 subunit of Pol II, and target Pol II for degradation (39, 40), we determined the Pol II response in VHL cells treated with Et743. Figure 4A shows that VHL-deficient (786-0) cells were deficient for Pol II degradation and instead accumulated hyperphosphorylated Pol II (Pol IIo). By contrast, the VHL-complemented cells behaved like other cell lines (proficient for TC-NER) and degraded Pol II in response to Et743. Those experiments suggest that VHL is implicated in the pathway leading to degradation of hyperphosphorylated Pol II in Et743-treated cells.

Next, we determined the effect of VHL on cell survival in response to Et743. Figure 4B (left) shows that VHL-complemented cells (VHL-C) were more sensitive to Et743 than their VHL-defective counterpart (VHL). To determine whether the effect of VHL on cell death was related to Pol II degradation, we treated the VHL-C cells with the proteasome inhibitor MG132 together with Et743. Figure 4B (right) shows that MG132 protected against Et743-induced cell killing. Together, these results indicate that VHL is involved in the degradation of Pol II in cells treated with Et743. They also suggest that Pol II degradation contributes to the antiproliferative activity of Et743.

Et743 is a potent transcription inhibitor (10–14) not only because it forms DNA adducts (15, 41) but also and most importantly because those adducts trap the TC-NER complex as it attempts to repair the Et743 adducts at sites where those adducts have arrested Pol II (17–19). In the present study, we show that Et743 also induces the degradation of Pol II. We provide evidence that Pol II degradation primarily affects elongating Pol II because proteasome inhibition leads to accumulation of hyperphosphorylated Pol II and that cells that fail to degrade Pol II accumulate its hyperphosphorylated form. We also provide evidence that Pol II degradation is TC-NER dependent based on lack of Pol II degradation in CSB and XP cells and on restoration of Pol II degradation on complementation of CSB and XPD cells.

Transcription elongation probably does not occur in a smooth continuous process even under normal conditions (23). Pol II is physiologically assisted by multiple factors that contend with chromatin structure (e.g., histone acetylases and methyltransferases enable chromatin relaxation). Transcription elongation is also contingent of damages that alter the DNA template. Damaged DNA can enter the active Pol II site and become buried deep inside the Pol II complex (42). TC-NER involves the recruitment of chromatin and DNA repair factors to the site where elongating Pol II is stalled, and enables the excision of a segment of ∼30 nucleotides surrounding the DNA adduct. Pol II can then be recycled for new rounds of transcription (23, 25, 38). Irreversible and prolonged transcription arrest has occasionally been linked to Pol II degradation in the case of UV-induced, cisplatin-induced, and 4-NQO–induced DNA adducts (22, 43, 44) and hydrogen peroxide–induced oxidative DNA lesions (45).

We found that Pol II degradation in response to Et743 is remarkably rapid and extensive in a wide range of cell lines (human colorectal carcinoma, Ewing's sarcoma, and fibroblasts). More than half of the total Pol II was lost within 1 h at pharmacologic (nanomolar) Et743 concentrations. Such characteristics make Et743 a potentially useful biochemical reagent to study the pathways involved in DNA damage-induced Pol II degradation. We focused on the TC-NER pathway because Et743 is known to affect TC-NER (17–19). We found that CSB cells were deficient, whereas XPC cells were proficient in degrading Pol II. Moreover, complementation of CSB cells normalized Pol II degradation. Because XPC is specifically involved in global genome DNA repair (GG-NER), which occurs in silent regions of the genome and in the nontranscribed strand of active gene, whereas CSB is required for TC-NER (24, 25, 38), we conclude that Pol II degradation is dependent on TC-NER (Fig. 5). Further support for the implication of TC-NER is based on the differential response of XPD-deficient and XPD-complemented cells, which showed defective and normal Pol II degradation, respectively, and on defective degradation of Pol II in XPA, XPF, and XPG cells.

A recent study in yeast showed that Pol II degradation in response to UV-induced DNA damage requires not only the CSB ortholog Rad26 but also an E3 ligase complex composed of elongins (Ela1 and Elc1), cullin 3 (Cul3), and Roc1 (Rbx1; ref. 22). Ribar et al. (22) proposed that in humans, the VHL ubiquitin ligase complex, composed of elongins B and C, cullin 2, and Rbx1, could act in an analogous manner. Our data obtained before that publication (22) agree with the possibility that the VHL complex could act in promoting Pol II degradation in DNA-damaged cells (Fig. 5). Such an activity extends the well-established role of VHL as an ubiquitin ligase that promotes the degradation of the hypoxia-inducible transcription factor-α (46). VHL has also recently been linked to multiple essential molecular pathways related to its tumor suppressor function (47). In addition, VHL-dependent degradation of Pol II has been reported in response to UV and cisplatin (39, 40), which, like Et743, elicit TC-NER repair (48, 49).

Finally, we find that VHL- and proteasome-dependent Pol II degradation is correlated with cell death, which is plausible because persistent Pol II down-regulation and lack of transcription would preclude the synthesis of essential cellular proteins. Because Et743 is remarkably active in soft tissue sarcoma and Ewing's sarcoma cells, it remains to be determined whether sarcoma cells have different specificity for TC-NER and Pol II degradation (35–37). Our study suggests that Pol II degradation could also play a therapeutic and prognostic role in the antitumor activity of Et743 and be tested as a biomarker of response in patients treated with Et743.

No potential conflicts of interest were disclosed.