Inhibition of nucleolar transcription by oxaliplatin involves ATM/ATR kinase signaling

SUMMARY

Platinum (Pt) compounds are an important class of anti-cancer therapeutics, but outstanding questions remain regarding their mechanism of action. Here, we demonstrate that oxaliplatin, a Pt drug used to treat colorectal cancer, inhibits rRNA transcription through ATM and ATR signaling, and induces DNA damage and nucleolar disruption. We show that oxaliplatin causes nucleolar accumulation of the nucleolar DNA damage response proteins (n-DDR) NBS1 and TOPBP1, however transcriptional inhibition does not depend upon NBS1 or TOPBP1, nor does oxaliplatin induce substantial amounts of nucleolar DNA damage, distinguishing the nucleolar response from previously characterized n-DDR pathways. Taken together, our work indicates that oxaliplatin induces a distinct ATM and ATR signaling pathway that functions to inhibit Pol I transcription in the absence of direct nucleolar DNA damage, demonstrating how nucleolar stress and transcriptional silencing can be linked to DNA damage signaling and highlighting an important mechanism of Pt drug cytotoxicity.

Graphical Abstract

eTOC Blurb

Oxaliplatin is an important chemotherapeutic known to induce nucleolar stress and inhibit RNA Polymerase I, distinct from the related Pt drug cisplatin. Nechay et al. demonstrate that inhibition of nucleolar transcription by oxaliplatin involves DNA damage signaling through ATM/ATR kinases in the absence of nucleolar DNA damage.

INTRODUCTION

Platinum (Pt) compounds have been a major cornerstone in cancer treatment for nearly 50 years. Cisplatin, the first Pt compound to receive FDA approval in 1978, is the standard of care for the treatment of bladder, lung, and germ cell cancers, among others¹. To overcome cellular resistance and reduce toxicity associated with cisplatin, further research has resulted in FDA approval of two additional Pt compounds to date – carboplatin and oxaliplatin. Despite the pervasiveness of Pt drugs in cancer treatment, challenges persist in defining their molecular mechanisms of action and predicting clinical outcomes². For example, colorectal and gastrointestinal cancers were long considered to be insensitive to Pt-based therapeutics (i.e. cisplatin and carboplatin), until oxaliplatin was discovered to be an effective treatment³. Thus, a better understanding of the relationship between structure and biological activity of Pt compounds is necessary to improve development and clinical application of this important family of therapeutics.

Most Pt-based drugs contain an electrophilic Pt(II) core which exhibits broad reactivity towards multiple classes of cellular macromolecules (e.g. DNA, RNA, proteins)^4,5. However, the prevailing hypothesis for Pt-induced cytotoxicity accounts primarily for the accumulation of Pt adducts on DNA, which leads to activation of the DNA damage response (DDR) and pro-apoptotic pathways⁶. This model, based primarily on studies using the first-generation drug cisplatin, is supported by a wealth of cellular and biochemical data as well as clinical evidence demonstrating that cancers deficient in certain DNA repair factors are sensitiized to Pt-based treatments⁷. While DNA lesions appear to be the primary mediator of cisplatin cytotoxicity, emerging work has illuminated diverse cellular mechanisms exhibited by Pt compounds containing different substituents around the Pt(II) core^8,9. In particular, phenotypic studies of oxaliplatin have indicated that this compound is better characterized as a transcription/translation inhibitor rather than a canonical DNA damaging agent¹⁰, which has led to investigation of alternative targets to explain its mechanism of action. Further, oxaliplatin disrupts nucleolar morphology and ribosome biogenesis^10–13, and reacts in vitro with nucleolar proteins¹², but the mechanism mediating nucleolar disruption and transcriptional arrest, and the extent to which different Pt compounds can evoke this response, is poorly understood.

Multiple cellular insults can lead to ‘nucleolar stress’, which is characterized by changes in nucleolar organization, impairment of ribosome biogenesis, and activation of pro-apoptotic signaling¹⁴. One major pathway that induces nucleolar disruption is the nucleolar DNA damage response (n-DDR), a subset of DDR signaling responding to ribosomal DNA (rDNA) damage within the nucleolus¹⁵. Selective induction of double-strand breaks (DSBs) in rDNA inhibits ribosomal RNA (rRNA) synthesis and leads to nucleolar segregation dependent on DDR factors such as ATM (Ataxia telangiectasia mutated) and ATR (Ataxia telangiectasia and Rad3-related) kinases, which signal through nucleolar adaptor proteins, such as the ATM-associated phosphoprotein Treacle. Interestingly, rRNA silencing and nucleolar reorganization can also result from DSBs occurring outside of nucleoli¹⁶, but the mechanism of in-trans damage signaling is not well characterized, and the propensity of different DNA damaging agents to induce this effect is unknown. In light of the connection between DNA damage signaling and Pt drugs, and the nucleolar stress induction by oxaliplatin, an examination of the role of the DDR in oxaliplatin-induced nucleolar disruption and cytotoxicity is warranted.

Here we characterize the mechanism of action of oxaliplatin in human cancer cell lines. Consistent with previous data^11–13,17, we find that oxaliplatin, but not cisplatin, is a potent and specific inhibitor of RNA Pol I transcription and rapidly induces nucleolar stress at clinically relevant concentrations. In contrast to proposed models of oxaliplatin-mediated nucleolar disruption and Pol I inhibition that invoke direct interaction of oxaliplatin with nucleolar components^10,12, we demonstrate that oxaliplatin-induced rRNA silencing involves ATM and ATR signaling that coincides with nuclear γH2AX foci and n-DDR factor accumulation in nucleoli, but in the absence of appreciable nucleolar DNA damage. While we observe the recruitment of established n-DDR factors NBS1 and TOPBP1, we show that these do not serve an essential role in transcriptional silencing by oxaliplatin. Our findings demonstrate an important contribution of a distinct ATM/ATR signaling pathway to oxaliplatin-induced silencing of nucleolar transcription. This pathway likely functions in concert with other complementary direct and indirect mechanisms of nucleolar disruption that together underly the diverse effects of oxaliplatin on cellular physiology.

RESULTS

Oxaliplatin is a selective inhibitor of nucleolar transcription

We first characterized the effect of oxaliplatin on nucleolar structure and rRNA transcription. Redistribution of nucleolar protein NPM1 from nucleoli to nucleoplasm is a common indicator of nucleolar stress¹⁸. Using NPM1 immunofluorescence, we observed predominantly nucleoplasmic distribution of NPM1 and nucleolar rounding in U2OS cells after 6 hr oxaliplatin treatment at near IC₅₀ drug concentration (Fig. 1A, SI Appendix, Fig. S1A, S1B). These morphological changes were also recapitulated in cells expressing mCherry-tagged NPM1 (SI Appendix, Fig. S1C), and we observed nucleolar caps containing Treacle (Fig. 1A)¹⁹. These perturbations to nucleoli are consistent with nucleolar stress induction by oxaliplatin, as reported previously^12,13,17.

Figure 1. Oxaliplatin selectively inhibits rRNA synthesis and induces nucleolar stress.

(A) Representative images of immunostaining against NPM1 (red) and Treacle (green) in U2OS cells with or without (±) treatment of oxaliplatin (5 μM) for 6 hours (3 independent replicates). (B) Schematic of “long-pulse” 5-EU labeling of RNA synthesis, and representative images of 5-EU incorporation in HeLa cells after treatment with Pt drugs (5 μM) or Act.D (5 nM) for 9 hours (2 independent replicates). White outlines of nuclei were generated from DAPI signal. Scalebars = 15 μM. (C) Schematic of 2’AzCyd metabolic labeling of RNA in U2OS cells co-treated with various drugs for 12 hours (left), and detection of extracted Cy3-labeled RNA by in-gel fluorescence (right top) and ethidium bromide staining (right bottom). (D) Quantification of RNA bands in Cy3 channel from (C), relative to signal in DMSO control. Error bars represent mean ± SEM of N = 3 biological replicates.

Previous work has shown close links between rRNA synthesis/processing, nucleolar stress and NPM1 redistribution¹⁸. To investigate, we used 5-ethynyluridine (5-EU) to label newly transcribed RNA during drug treatment (SI Appendix, Fig. S1D, S1E). Compared to control, oxaliplatin-treated cells exhibited specific depletion of 5-EU signal in nucleoli (Fig. 1B, SI Appendix, Fig. S1F). A similar pattern was observed with Actinomycin D (Act.D), a selective Pol I inhibitor, suggesting that oxaliplatin selectively inhibits rRNA synthesis rather than global transcription. In contrast, 5-EU signal in cisplatin-treated cells was indistinguishable from control (Fig. 1B). To confirm, we metabolically labeled RNA with 2’-azidocytidine (2’-AzCyd)²⁰, and visualized 18S and 28S rRNA bands by in-gel fluorescence during drug treatment (Fig. 1C). Oxaliplatin or Act.D treatment nearly abolished signal from nascent 18S and 28S rRNA (Fig. 1C and 1D), while cisplatin induced only a minimal decrease in rRNA transcription, consistent with 5-EU labeling results. Together, these data show that oxaliplatin strongly inhibits rRNA synthesis and induces nucleolar stress, distinct from cisplatin.

To investigate the relationship between nucleolar disruption and rRNA transcription, we monitored rRNA synthesis and NPM1 localization in oxaliplatin-treated cells using 5-EU labeling and immunofluorescence (Fig. 2A, 2B). At 5 μM oxaliplatin treatment, nucleolar transcription and NPM1 localization were perturbed by ~80% and ~50%, respectively, compared to untreated cells. Surprisingly, nucleoplasmic redistribution of NPM1 and nucleolar transcription inhibition were detected with oxaliplatin concentrations nearly 5-fold lower than IC₅₀ (SI Appendix, Fig. S1A, S1B). By comparison, cisplatin induces nucleolar stress only at dosages 10-fold above IC₅₀, a level not considered clinically relevant¹¹. Next, we characterized the timing of oxaliplatin-induced nucleolar stress and transcriptional inhibition, and measured nucleolar cap formation by Treacle in addition to NPM1 distribution. We observed NPM1 dispersion and Treacle cap formation beginning concurrently as early as 2 hr of oxaliplatin treatment, and reaching steady state within 4 hr (Fig. 2C). We compared this to the kinetics of nucleolar transcription inhibition, which also showed substantial inhibition at 2 hr (Fig. 2D). The similarities in kinetics suggest that Pol I inhibition by oxaliplatin is concurrent with or even slightly precedes nucleolar reorganization.

Figure 2. Oxaliplatin is a potent and rapid inhibitor of nucleolar function.

(A) Schematic of “short-pulse” 5-EU labeling of nucleolar RNA synthesis (top), and representative images of NPM1 localization and rRNA silencing after 6 hr oxaliplatin treatment (bottom). Scalebars = 15 μM. White outlines are nuclei from DAPI signal. (B) Quantification of 5-EU signal in nuclei from (A), normalized relative to DMSO control (left). Quantification of NPM1 localization from (A) as measured by coefficient of variation (CV) in NPM1 signal in nuclei (right). Number of cells analyzed for each condition, summed over 2 biological replicates: DMSO (N = 103), oxaliplatin at 0.625 μM (N = 121), 1.25 μM (N = 145), 2.5 μM (N = 136), and 5.0 μM (N = 104). (C) Quantification of Treacle (TCOF1) foci count per nucleus and NPM1 localization by CV in response to oxaliplatin (5 μM) over 0–5 hr. Number of cells analyzed for each condition, summed over 3 biological replicates: 0 hours (N = 129), 1 hour (N = 257), 2 hours (N = 111), 3 hours (N = 141), 4 hours (N = 269), 5 hours (N = 119). (D) Quantification of rRNA synthesis by nucleolar 5-EU incorporation in oxaliplatin-treated cells (5 μM) over 0–5 hours. Number of cells analyzed for each condition, summed over 2 biological replicates: 0 hours (N = 153), 1 hour (N = 102), 2 hours (N = 172), 3 hours (N = 136), 4 hours (N = 137), 5 hours (N = 96). (E) Western blot of p53 stabilization after 10 μM oxaliplatin and cisplatin treatments, with actin loading control underneath (top). Quantification of p53 signal from western blot (bottom). Error bars represent mean ± SEM of N = 2 biological replicates. Statistical analysis in (B-E) was performed using a one-way analysis of variance, Tukey’s multiple comparisons test. Adjusted p-values are indicated by ***P<0.001, **P<0.01, *P<0.05, and n.s. denotes a non-significant p-value > 0.05. All experiments were performed in wild-type U2OS cells.

Nucleolar stress can induce p53-dependent apoptosis^18,21. Previous work found no significant difference in p53 stabilization induced by prolonged cisplatin or oxaliplatin treatment¹⁷. We measured p53 levels and indeed saw similar increases in p53 after 24 hr cisplatin or oxaliplatin treatment (Fig. 2E). However, we observed 2-fold higher p53 levels in oxaliplatin-treated cells compared to cisplatin at 6 hr, shortly after nucleolar stress induction. This early apoptotic signaling following nucleolar stress suggests a direct connection between nucleolar stress and oxaliplatin cytotoxicity.

Oxaliplatin binds nucleolar nucleic acids and proteins with lower efficiency than cisplatin.

Previous studies have demonstrated that ribosomal DNA damage can lead to Pol I inhibition and redistribution of nucleolar proteins^16,22,23. Given the affinity of Pt ions for guanine-rich nucleic acids^5,24, we speculated that oxaliplatin may form bulky adducts on rRNA/rDNA that disrupt interactions with nucleolar proteins. We first analyzed Pt accumulation on rRNA using primer extension analysis on RNA isolated from Pt-treated U2OS cells²⁵. We probed a solvent-accessible region on the 28S rRNA adjacent to the peptidyl transferase center (PTC) containing known reactive sites for cisplatin and ribosomal inhibitors^26,27. While we measured several dose-dependent cisplatin adducts in the vicinity of the E site, we were unable to observe oxaliplatin adducts in this region (Fig. 3A). Primer extension probing of 18S and 5S rRNA showed a similar trend with relatively little accumulation of oxaliplatin compared with cisplatin (SI Appendix, Fig. S2A, S2B), suggesting that rRNA is not a major target of oxaliplatin.

Figure 3. Oxaliplatin does not preferentially target nucleolar nucleic acids and transcription factors.

(A) Schematic of primer extension assay for Pt adducts on cellular RNA (top). Primer extension of 28S human rRNA extracted from U2OS cells treated with 0–100 μM of cisplatin or oxaliplatin for 12 hr (3 biological replicates). Dideoxy sequencing ladder is shown on left. Annotations of local rRNA structures are provided on right. (B) Western blot of total cell extract from U2OS cells treated with Act. D (10 nM), BMH-21 (1 μM), oxaliplatin (10 μM), or cisplatin (10 μM) for 6 hr. Depletion of RPA194, and to a lesser extent TAF I p110, relative to actin control is seen only for BMH-21 treated cells (2 biological replicates). (C) Representative images of immunostaining against RPA194 (red) and Treacle (green) in cells treated as in (B), from 3 biological replicates. Scalebars = 15 μM. (D) Gel analysis of in vitro Pt binding to rDNA promoter after incubation with cisplatin or oxaliplatin for 12 hr at 37 °C in PtNA buffer (10 mM Na₂PO₄, 100 mM NaNO₃, 1 mM Mg(NO₃)₂). 2 biological replicates performed. (E) Schematic of RNA Pol I pre-initiation complex (PIC) pulldown from extracted U2OS nuclear lysate using streptavidin bead-immobilized DNA oligo containing rDNA promoter sequence, with (I) or without (II) 0.3 M KCl washing (top). Western blot analysis of RPA194 and UBF after incubation (1 hr, 4 °C) with rDNA promoter in cell lysate containing DMSO (1% v/v), 9-hydroxyellipticine (40 μM), Act.D (500 nM), or oxaliplatin (100 μM). 2 biological replicates (bottom, left). Western blot analysis of recovered UBF, RPA194, and NPM1 after incubation in cell lysate with Act. D (500 nM), cisplatin (100 μM), or oxaliplatin (100 μM) for 1 hr at RT, followed by washes with 0.3 M KCl (2 biological replicates) (bottom, right).

We next investigated the rDNA promoter region, which is crucial for transcription initiation and has been identified as a target for Pol I inhibitors²⁸ such as BMH-21, which blocks promoter escape and transcription elongation leading to proteasome-mediated degradation of the large subunit (RPA194)²⁹. We did not observe depletion of RPA194 in oxaliplatin-treated cells, as measured by western blot (Fig. 3B) or immunofluorescence (Fig. 3C, SI Appendix, Fig. S2C), indicating that oxaliplatin has a distinct mechanism of action from BMH-21. Instead, oxaliplatin treatment results in RPA194 accumulation into nucleolar caps, in agreement with previous studies of Pol I redistribution in response to transcriptional silencing³⁰.

Another potential mechanism of Pol I silencing is through inhibition of pre-initiation complex (PIC) assembly at the rDNA promoter^31,32. We assayed the general reactivity of Pt compounds to DNA by treating a 400-bp rDNA segment containing the promoter sequence with Pt drug for 12 hr and analyzing adducts by dPAGE. We observed substantially greater amounts of Pt-DNA adducts with 10–50 μM cisplatin as compared to 10–50 μM oxaliplatin (Fig. 3D), consistent with prior reports demonstrating greater reactivity of cisplatin towards nucleic acids²⁴. Next, we captured pre-initiation complex from U2OS nuclear extract using a biotinylated DNA fragment³³. In the presence of 100 μM oxaliplatin, we found that promoter binding of Pol I and Upstream Binding Factor (UBF) was unaffected (Fig. 3E, left). To test potential crosslinking of rDNA and proteins by oxaliplatin, we used high salt washes to remove non-covalently bound PIC components. Under these conditions, we observed retention of Pol I, UBF, and NPM1 upon 100 μM oxaliplatin treatment (Fig. 3E, right), indicative of covalent crosslinking. However, 100 μM cisplatin treatment resulted in similar or greater amounts of crosslinking with all three proteins. We conclude that interactions with rDNA promoter and nucleolar proteins are unlikely to fully explain the specific effect of oxaliplatin on rRNA transcription. Further, in vitro covalent crosslinking with high concentrations of either Pt compound does not mean that therapeutically relevant dosages can crosslink rDNA and nucleolar proteins in cells.

Nevertheless, based on our findings, we investigated whether reactivity with nucleolar proteins could underlie transcriptional inhibition in cells. Recent work has shown that oxaliplatin can react in vitro with fibrillarin (FBL) and proposed that modification of FBL or other nucleolar components is involved in oxaliplatin-mediated nucleolar disruption¹². To test this hypothesis, we overexpressed FBL and measured impact on oxaliplatin activity, since overexpression was previously shown to confer resistance¹². While we observed a modest increase (<2-fold) in IC₅₀ upon FBL overexpression (SI Appendix, Fig. S2F), we observed no difference in nucleolar transcription (SI Appendix, Fig. S2D, S2E). This result suggests that modification of FBL by oxaliplatin is unlikely to be a key driver for Pol I inhibition. In sum, our data indicates that transcriptional inhibition by oxaliplatin is unlikely to be caused by direct interactions with nucleolar components, making its mechanism of action distinct from other major classes of Pol I inhibitors.

Transcriptional silencing by oxaliplatin involves ATM and ATR activity

Since Pt(II) drugs can activate DDR in cells, we explored whether oxaliplatin-DNA lesions might contribute to Pol I inhibition. Notably, studies have shown that global DNA damage or nucleolar DNA damage can directly lead to transcriptional silencing^15,34,35. In particular, global silencing of nucleolar transcription can occur in response to DNA damage localized outside of the nucleolus by a mechanism involving ATM and other DDR proteins¹⁶. Although recent studies have questioned DDR activation as a major mediator of oxaliplatin-induced nucleolar stress^10,17, the existence of specific DDR pathways communicating to nucleoli prompted us to investigate the relationship of oxaliplatin to DNA damage signaling.

First, we assayed a panel of Pt compounds and Pol I inhibitors for the formation of γH2AX foci, a marker for DNA damage, in parallel with 5-EU incorporation (Fig. 4A). We observed γH2AX foci within 6 hr of treatment with either cisplatin and oxaliplatin, though in agreement with previous reports, oxaliplatin-treated cells exhibited less γH2AX signal (Fig. 4B). Interestingly, the Pol I inhibitor CX-5461 induced a greater number of γH2AX foci than oxaliplatin, consistent with a recent study reclassifying it as a DNA-damaging topoisomerase poison³⁶. As expected, Act.D generated few γH2AX foci at the low dosage used for selective inhibition of Pol I³⁷. We conducted a time-course analysis to determine whether DNA damage precedes transcriptional inhibition (SI Appendix, Fig. S3A, S3B) and found that these processes are concurrent. We conclude that although the extent of DNA damage does not appear to correlate strongly with Pol I inhibition, oxaliplatin is capable of activating DDR pathways together with nucleolar stress induction.

Figure 4. rRNA silencing by oxaliplatin is dependent on ATM and ATR activity.

(A) Representative images of γH2AX immunostaining and 5-EU labeling after treatment of U2OS cells with cisplatin (5 μM), oxaliplatin (5 μM), Act.D (5 nM), or CX-5461 (1 μM) for 6 hr. (B) Quantification of changes in nucleolar 5-EU incorporation in response to drug treatments, relative to DMSO control (top), and quantification of γH2AX foci/nucleus, measured in ImageJ using ROIs from DAPI channel (bottom). Number of cells analyzed for each condition, summed over 3 biological replicates: DMSO (N = 151), cisplatin (N = 135), oxaliplatin (N = 142), Act.D (N = 121), CX-5461 (N = 164). Error bars represent mean ± SEM. (C) Representative images of 5-EU incorporation and Treacle localization into nucleolar caps upon co-treatment of oxaliplatin (5 μM) with or without VE-821 (ATRi) and KU-55933 (ATMi) for 6 hr. (D) Quantification of 5-EU signal in nuclei from (C), normalized relative to DMSO control (top). Quantification of Treacle foci/nucleus from (C) (bottom). Number of cells analyzed for each condition, summed over 3 biological replicates: DMSO (N = 141), oxaliplatin (N = 142), with ATRi (N = 143), with ATMi (N = 176), with both (N = 155). (E) Quantification of 5-EU signal in nuclei of cells co-treated with oxaliplatin (5 μM), CX-5471 (1 μM), or Actinomycin D (5 nM), with or without ATRi and ATMi, for 6 hr. At least N = 50 cells analyzed for each condition over 2 biological replicates. (F) Representative images of rRNA synthesis measured by 5-EU incorporation after co-treatment with oxaliplatin (5 μM) and inhibitors for CHK1 (300 nM), CHK2, (300 nM), or both CHK1/CHK2 (10 μM) (left), and quantification of 5-EU incorporation, relative to DMSO control (right). Number of cells analyzed in each condition, summed over 2 biological replicates: DMSO (N = 143), OxaPt/mock (N = 118), +CHK1i (N = 160), +CHK2i (N = 127), +CHK1/2i (N = 141). Statistical analysis in (A-F) was performed using a one-way analysis of variance, Tukey’s multiple comparisons test. Adjusted p-values are indicated by ***P<0.001, *P<0.05, and n.s. denotes a non-significant p-value > 0.05. All experiments were performed in wild-type U2OS cells. In (A), (C), and (F), white outlines of nuclei were generated from the DAPI channel. Scalebars = 15 μM.

Next, we explored whether canonical DDR signaling pathways are relevant to oxaliplatin-mediated transcriptional inhibition. We inhibited ATM and ATR kinases, which phosphorylate many DDR proteins³⁸, with small molecules (KU-59933 and VE-821, respectively) and measured effects on 5-EU incorporation and nucleolar structure during oxaliplatin treatment (Fig. 4C). Gratifyingly, inhibition of either ATM or ATR resulted in partial rescue of transcription, while inhibition of both kinases restored nucleolar transcription to ~60% of control (Fig. 4D). Treatment of cells with ATM/ATR inhibitor alone or in combination did not affect nucleolar transcription (SI Appendix, Fig. S3C, S3D). We observed similar effects in HCT116 and A549 cells (SI Appendix, Fig. S3E–S3H), indicating that the phenomenon was general across diverse cancer cell lines, but the magnitude of rescue varied. Interestingly, inhibition of ATM (or both ATM and ATR) resulted in near complete restoration of nucleolar transcription in oxaliplatin-treated HCT116, whereas A549 cells showed more modest response, and greater rescue with ATR inhibition. The formation of Treacle-containing nucleolar caps (Fig. 4D) and the nucleoplasmic redistribution of NPM1 (SI Appendix, Fig. S4A, S4B) was also partially abrogated by these inhibitors, demonstrating that ATM and ATR activity contributes to nucleolar segregation, although rescue of Pol I transcription was more robust than restoration of nucleolar morphology.

We next asked whether other Pol I inhibitors utilize ATM/ATR signaling, but we did not observe any transcriptional rescue for either Act.D or CX-5461 (Fig. 4E). Thus, oxaliplatin-induced silencing of rRNA transcription is mechanistically distinct from Act.D or CX-5461 and depends upon activation of ATM and ATR signaling. Further, we investigated downstream targets of ATM/ATR. Transcription was partially rescued by inhibition of Chk1/Chk2 (Fig. 4F), albeit to a lesser extent than ATM/ATR inhibition. Importantly, treatment with oxaliplatin resulted in Chk1/Chk2 phosphorylation, indicating bonafide activation of ATM/ATR signaling (SI Appendix, Fig. S4C, S4D). Further, oxaliplatin-mediated p53 stabilization depends in part upon ATM/ATR signaling (SI Appendix, Fig. S4E–S4G). Taken together, these data indicate that oxaliplatin-induced rRNA silencing depends in large part upon ATM/ATR signaling and also implicate kinase activity in nucleolar disruption and p53 induction.

The nucleolar response to oxaliplatin shares features with n-DDR signaling

We next characterized the mechanism bridging DNA damage signaling with transcriptional inhibition. We considered the possibility that a nucleolar DNA damage response (n-DDR) is operative. Previous studies of n-DDR have targeted nucleolar DSBs with IR or I-Ppo1 endonuclease and characterized a transcriptional silencing mechanism centered around Treacle-mediated recruitment of DDR factors, including NBS1^15,16,39,40. We found that 31% of oxaliplatin-treated cells contain NBS1 perinucleolar foci co-localizing with UBF-positive nucleolar caps (Fig. 5A). Recruitment of NBS1 started at the interior of the nucleoli prior to segregation into peripheral cap structures (SI Appendix, Fig. S5A), consistent with a previous report using I-Ppo1⁴¹, and was dependent upon Treacle (Fig. 5B). Notably, minimal NBS1 recruitment was observed after Act.D or doxorubicin treatment and only 15% of CX-5461 treated cells contained NBS1-positive caps (Fig. 5C). In contrast, I-Ppo1 produced nucleolar NBS1 in 87% of transfected cells. Given that doxorubicin and CX-5461 are potent DNA damaging agents, these results indicate that Treacle-mediated recruitment of NBS1 is specific to certain forms of DNA damage that activate the n-DDR response.

Figure 5. Oxaliplatin induces nucleolar recruitment of NBS1 but not of other repair-associated DDR factors.

(A) Representative images of NBS (red) and UBF (green) immunostaining after oxaliplatin treatment. White arrows indicate NBS1 colocalization with UBF-positive nucleolar caps. Number in bottom right indicates percentage of cells exhibiting nucleolar caps containing NBS1; quantified in C from 3 biological replicates. (B) Fraction of I-Ppo1 or oxaliplatin-treated cells containing NBS1-positive nucleolar caps with or without siRNA knockdown of endogenous Treacle. Error bars represent mean ± SEM of 2 biological replicates. (C-D) Quantification of cells containing NBS1-positive nucleolar caps, as indicated by peripheral UBF foci around nucleoli, after treatment with indicated drugs or I-Ppo1. Error bars represent mean ± SEM of 3 biological replicates. (E) Fraction of I-Ppo1 or oxaliplatin-treated cells containing NBS1-positive nucleolar caps with or without siRNA knockdown of endogenous TOPBP1. Error bars represent mean ± SEM of 2 biological replicates. Statistical analysis in (B-E) was performed using a one-way analysis of variance, Tukey’s multiple comparisons test. Adjusted p-values are indicated by ***P<0.001, *P<0.05, and n.s. denotes a non-significant p-value > 0.05. (F-G) Representative images of MRE11 and NBS1 immunoblotting (F) or UBF and phospho-RPA2 (G) after treatment with I-Ppo1 or oxaliplatin (2 independent replicates). All experiments performed in U2OS cells. Scalebars = 20 μM.

Treacle-mediated NBS1 recruitment and nucleolar segregation during n-DDR depend on ATM and ATR signaling^15,34,41. Consistent with previous findings, I-Ppo1 expressing cells treated with ATM or ATR inhibitors possessed fewer NBS1-positive nucleolar caps (Fig. 5D). Similarly, NBS1 recruitment in oxaliplatin cells was sensitive to ATR or ATM inhibition. NBS1 recruitment after I-Ppo1 expression also depends on TOPBP1, an ATR activator proposed to interact directly with nucleolar Treacle⁴¹. Interestingly, we observed no change in NBS1 recruitment in oxaliplatin cells upon TOPBP1 knockdown, while I-Ppo1 transfected cells showed TOPBP1-dependent NBS1 localization (Fig. 5E). Interestingly, TOPBP1 still accumulated in nucleolar caps in a subset of oxaliplatin-treated cells, despite apparent lack of interaction with NBS1 (SI Appendix, Fig. S5B).

Since NBS1 and TOPBP1 foci were only observed in some oxaliplatin-treated cells, we investigated whether their role is linked to cell cycle state. Since the checkpoint response to DNA damage is typically activated by replication stress⁴², we established that γH2AX activation by oxaliplatin occurs primarily in S-phase cells by EdU incorporation (SI Appendix, Fig. S5C). We further found strong cell cycle dependency of nucleolar NBS1 recruitment, with ~60% of EdU-positive cells exhibiting nucleolar NBS1 foci as compared with less than 10% of EdU-negative cells (SI Appendix, Fig. S5D). We also tested whether transcriptional recovery by ATM/ATR inhibition was cell-cycle specific by combining our 5-EU labeling assay with Cyclin A staining (SI Appendix, Fig. S5E, S5F). We observed transcriptional rescue in both G1 and S/G2 cells, but rescue was slightly more efficient in S/G2 cells. Notably, even in non-S-phase cells lacking observable γH2AX foci and nucleolar NBS1, we still observe signs of nucleolar disruption. We conclude that oxaliplatin activates DDR pathways involving NBS1 in S-phase cells, but this pathway is unlikely to be the universal mechanism of nucleolar stress induction. Thus, the nucleolar response to oxaliplatin appears to be distinct from canonical n-DDR mechanisms for rDNA damage, despite sharing a few similarities.

NBS1 and TOPBP1 are not essential for oxaliplatin-mediated rRNA synthesis inhibition

Since the nucleolar response to oxaliplatin differs from I-Ppo1, we explored alternative n-DDR pathway which respond to DSBs outside of the nucleolus. Larsen et al.¹⁶ reported an in-trans n-DDR mechanism distinguished by lack of MRE11 recruitment to nucleoli, indicating absence of active DNA repair in rDNA loci. Moreover, individual nucleoli silenced by this pathway also lack γH2AX, in contrast to direct nucleolar rDNA damage^40,41. To investigate an in-trans n-DDR, we quantified γH2AX foci near nucleoli. While most oxaliplatin-induced γH2AX foci were distributed throughout the nucleus, a minor fraction of cells (~6%) exhibited γH2AX foci overlapping with Treacle-positive nucleolar caps (SI Appendix, Fig. S6A–S6C), however, they were considerably less intense and frequent than in I-Ppo1 expressing cells and were also observed at similar frequency after cisplatin treatment (SI Appendix, Fig. S6C). We also evaluated nucleolar R loops after oxaliplatin treatment and found reduced formation, likely due to transcriptional inhibition; a similar result was obtained upon Act.D treatment, while CX-4561 increased nucleolar R loops (SI Appendix, Fig. S6D, S6E). Furthermore, we did not observe any colocalization of endogenous MRE11 with NBS1 foci in oxaliplatin-treated cells (Fig. 5F, SI Appendix, Fig. S6F), nor did we observe phosphorylated RPA2 (RPA2 pS4/S8) (Fig. 5G, SI Appendix, Fig. S6G), suggesting that oxaliplatin does not induce active DNA repair. In contrast, I-Ppo1 transfection resulted in both MRE11 and p-RPA2 within nucleolar caps (Fig. 5F and Fig. 5G). Based on the absence of repair factors and minimal γH2AX signaling within nucleoli, we conclude that oxaliplatin-mediated nucleolar disruption is not a result of direct damage to ribosomal DNA, but rather due to in-trans signaling.

Although nucleolar recruitment of NBS1 is not a general feature of oxaliplatin treatment, all reported n-DDR mechanisms have proposed a role for NBS1 in rRNA silencing. We therefore utilized a U2OS cell line expressing a hypomorphic C-terminal fragment of NBS1, hereafter referred to as “NBS1ΔN”, which is deficient in DNA end resection and does not undergo nucleolar segregation or rRNA silencing in response to rDNA damage by I-Ppo1^41,43. When we treated NBS1ΔN cells with oxaliplatin, we observed an even greater decrease in rRNA transcription relative (Fig. 6A, 6B), while transcription upon I-Ppo1 expression was higher as previously reported. However, oxaliplatin-induced transcriptional inhibition was still partially reversed in NBS1ΔN cells by co-treatment of inhibitors for ATM and ATR (Fig. 6C). Furthermore, we did not observe differences in nucleolar transcription for oxaliplatin-treated cells lacking TOPBP1, whereas I-Ppo1 induced silencing was substantially reversed under these conditions (Fig. 6D, 6E). Together, these results demonstrate that although nucleolar recruitment of NBS1 and TOPBP1 is observed in a subset of oxaliplatin-treated cells, neither is essential for ATM/ATR signaling to nucleoli and subsequent transcriptional inhibition, in contrast to reported n-DDR mechanisms. Nonetheless, we determined that NBS1ΔN cells are >2-fold more sensitive to oxaliplatin, likely due to widespread DNA repair defects at oxaliplatin-DNA lesions (Fig. 6F). Similarly, combining oxaliplatin with ATR inhibition or ATR and ATM inhibition also resulted in modest sensitization (Fig. 6G). By comparison, sensitization of NBS1ΔN cells to cisplatin was observed to a lesser extent, suggesting that NBS1 plays an important role in the repair of oxaliplatin-induced lesions but does not mediate transcriptional inhibition and nucleolar disruption. Based on these findings, we propose that oxaliplatin inhibits rRNA transcription through ATM/ATR-dependent signaling that is independent of NBS1, and does not result from nucleolar DNA damage.

Figure 6. Full NBS1 function is not essential for rRNA silencing by oxaliplatin.

(A) Representative images of rRNA synthesis in wild-type and “NBS1ΔN” U2OS cells measured by 5-EU incorporation after treatment with oxaliplatin (5 μM, 6 hr) or transfection of I-Ppo1 (8 hr). (B) Quantification of 5-EU signal from (A), relative to DMSO control in WT cells. Number of cells analyzed in each condition, summed over 2 biological replicates: DMSO in WT (N = 143) and NBS1ΔN (N = 142), oxaliplatin in WT (N = 112) and NBS1ΔN (N = 116), and I-Ppo1 in WT (N = 67) and NBS1ΔN (N = 65). For I-Ppo1-treated cells, only cells positive for transfection by HA staining were quantified. (C) Quantification of 5-EU signal in cells co-treated with oxaliplatin and ATRi, ATMi, or both, relative to DMSO control. At least N = 100 cells analyzed for each condition over 2 biological replicates. (D) Measurement of rRNA synthesis by 5-EU incorporation after treatment with oxaliplatin (5 μM, 6 hr) or transfection with I-Ppo1 (8 hr) in siCtrl and siTOPBP1-treated cells. (E) Quantification of 5-EU incorporation from (A), relative to DMSO control. Number of cells analyzed in each condition, summed over 2 biological replicates: DMSO in siCtrl (N = 43) and siTOPBP1 (N = 72), OxaPt in siCtrl (N = 64), and siTOPBP1 (N = 59), I-Ppo1 in siCtrl (N = 46) and siTOPBP1 (N = 50). For I-Ppo1-treated cells, only cells positive for HA staining were quantified. (F) Dose response curves for oxaliplatin and cisplatin in wild-type and NBS1ΔN U2OS cells. Cell viability was measured using MTS assay at 72 hours. Error bars represent mean ± SEM of N = 6 biological replicates. (G) Dose response curves for oxaliplatin with or without VE-821 (ATRi, 4 μM) or KU-55933 (ATMi, 10 μM) in wild-type U2OS cells. Cell viability was measured using MTS assay at 72 hours. Error bars represent mean ± SEM of N = 3 biological replicates. Statistical analysis in (B), (C), and (E) was performed using a one-way analysis of variance, Tukey’s multiple comparisons test. Adjusted p-value is indicated by ***P<0.001, **P<0.01, and n.s. denotes a non-significant p-value > 0.05. White outlines of nuclei in (A) and (D) were generated from the DAPI channel. Scalebars = 15 μM.

DISCUSSION

Herein, we investigate cellular signaling pathways responsible for coordinating disruption of nucleolar organization and rRNA transcription induced by oxaliplatin. We show that oxaliplatin-induced rRNA silencing involves ATM-Chk2 and ATR-Chk1 pathways, and demonstrate that small-molecule inhibition of either pathway rescues nucleolar transcription in multiple cancer cell lines. Further, we find that oxaliplatin does not induce substantial nucleolar DNA damage, and that oxaliplatin-induced rRNA silencing is independent of NBS1 or TOPBP1, distinguishing it from previously characterized cis and trans n-DDR pathways. Our findings indicate that oxaliplatin induces a distinct ATM/ATR signaling pathway that functions to inhibit Pol I transcription, contributing to activation of p53 and disruption of ribosome biogenesis, and ultimately cell death. Our study demonstrates a previously underappreciated connection between oxaliplatin-induced nucleolar disruption and DNA damage signaling, and together with complementary studies^10,12,13,17 illuminates the diverse mechanisms of action exhibited by chemotherapeutic Pt drugs.

We demonstrate that nucleolar stress induction by oxaliplatin shares many characteristics with previously described n-DDR signaling pathways. In particular, lack of MRE11 and phospho-RPA2 in nucleoli is consistent with in-trans n-DDR signaling resulting from DNA damage outside of nucleoli¹⁶, and distinguishes oxaliplatin-induced DDR signaling from n-DDR responding to direct rDNA damage. Notably, while NBS1 and TOPBP1 accumulate in nucleolar caps dependent on Treacle, neither protein is required for transcriptional silencing, in contrast to reported in trans signaling¹⁶. Since NBS1/TOPBP1 accumulation and transcriptional silencing depend upon ATM/ATR activity, we posit that these pathways function in parallel. The precise mechanism of Pol I inhibition through ATM/ATR is not known. Some Pol I transcription factors are known ATM/ATR targets⁴⁴, and ATM-dependent histone modifications can affect rDNA transcription upon DNA damage⁴⁵. Given the mechanistic diversity in n-DDR signaling, further studies are necessary to characterize nucleolar ATM/ATR substrates upon DNA damage and cellular stress.

Oxaliplatin induces nucleolar NBS1 accumulation independent of MRE11, but little is known about NBS1 outside of the MRN complex. In oxaliplatin-treated cells, nucleolar NBS1 correlates strongly with γH2AX foci and replication stress in S-phase, suggesting a threshold level of DNA damage required for nucleolar accumulation. Although the absence of MRE11 and phospho-RPA2 suggests otherwise, we cannot rule out active DNA repair in nucleolar caps containing NBS1. However, given efficient rRNA silencing in cells lacking nucleolar NBS1 and γH2AX foci, we conclude that transcriptional silencing is independent of NBS1, or that multiple pathways facilitate nucleolar disruption. We show that nucleolar NBS1 recruitment is not seen upon Pol I inhibition by other DNA-damaging drugs like CX-5461 and doxorubicin, or the non-DNA damaging inhibitor Act.D. Importantly, NBS1-deficient cells are more sensitive to oxaliplatin, likely due to defective DNA repair – therefore, the potentially multifaceted role of NBS1 during oxaliplatin treatment warrants further investigation.

The cytotoxicity of Pt drugs is generally thought to arise from Pt-DNA adducts, yet our report and previous studies show that oxaliplatin generates less DNA damage than cisplatin, but is comparably cytotoxic⁴⁶. While researchers have proposed that oxaliplatin disrupts nucleoli and rRNA transcription independent of DNA damage, our study shows that ATM/ATR signaling is involved in rRNA silencing, and demonstrates DNA damage foci and n-DDR activation in oxaliplatin-treated cells. In contrast, cisplatin fails to induce nucleolar phenotypes except at high concentrations above clinically relevant dosages. How do these two structurally related Pt drugs generate such different outcomes? We propose that DDR signaling to oxaliplatin lesions is driven by specific recognition of oxaliplatin- adducts rather than by the total amount of generated damage. In support of this idea, biophysical analyses of cisplatin and oxaliplatin adducts on DNA have identified differences in bending angle, hydrogen bonding patterns, and conformational flexibility of the local DNA structure, all of which have been proposed to affect the affinity of damage recognition proteins for Pt adducts^47–49. In addition, several HMG-domain proteins, such as UBF and HMGB1, as well as mismatch repair protein MSH2, have been shown in vitro to bind cisplatin-DNA and oxaliplatin-DNA adducts with different affinities, which may result in differential recognition and DDR signaling in cells^50–52.

An alternative hypothesis to explain the unique ability of oxaliplatin among Pt drugs to induce nucleolar stress and inhibit nucleolar transcription is that oxaliplatin selectively accumulates in nucleoli and directly modifies nucleolar biomolecules. In this model, oxaliplatin directly modulates nucleolar structure/dynamics, which then activates ATM/ATR signaling to inhibit RNA Pol I^15,16,53, resulting in a feedback loop to further impair nucleolar assembly. Interestingly, recent work has highlighted the propensity for some antineoplastic drugs to preferentially partition into phase-separated molecular condensates including the nucleolus⁵⁴. It is currently unknown whether this behavior is driven by specific small molecule-condensate interactions or through general physicochemical properties such as molecular size and hydrophobicity, however the increased hydrophobicity of the oxaliplatin DACH ligand in comparison to polar cisplatin NH₃ ligands makes a partitioning mechanism worthy of further exploration. While our manuscript was in review, Schmidt et al.¹² reported that oxaliplatin can react in vitro with the nucleolar scaffolding protein FBL and modulate its ability to undergo phase separation, and proposed that the selective partitioning of oxaliplatin into nucleoli combined with modification of nucleolar proteins and RNA underlies its activity. Direct modulation of nucleolar structure by oxaliplatin modification may serve to explain why rescue of Pol I transcription by ATM/ATR inhibition is incomplete in A549 and U2OS cells, and also why restoration of nucleolar morphology (as measured through NPM1 localization) is less dramatic than transcriptional recovery. Nevertheless, there are several aspects of the oxaliplatin nucleolar accumulation/modification model that are inconsistent with data presented herein and independently published work and should be further investigated. First, we show by primer extension that cisplatin accumulates to a greater degree than oxaliplatin on 28S, 18S, and 5.8S rRNA sequences. Second, both cisplatin and oxaliplatin induce a similarly minor amount of nucleolar γH2AX foci, and we do not observe robust activation of in-cis n-DDR after oxaliplatin treatment. Finally, overexpression of FBL does not lead to restoration of nucleolar structure or rRNA synthesis, and only confers modest resistance to oxaliplatin treatment in U2OS cells, suggesting that FBL modification alone does not initiate nucleolar collapse. In addition, a recent study using click chemistry-compatible oxaliplatin- and cisplatin-like probes⁵⁵ showed that nucleolar accumulation and cellular distribution of diverse Pt compounds is largely similar and uncorrelated with their ability to induce nucleolar stress. Moving forward, further characterization of oxaliplatin lesions in the context of nucleolar structure and function will be necessary to identify the factors and likely multiple pathways that recognize and respond to oxaliplatin treatment in cancer cells. In addition, and as highlighted in this study, the disruption of ribosome biogenesis through specific DNA damage signaling pathways is a promising therapeutic strategy and may inform development of the next generation of Pol I transcription inhibitors.

Limitations of this study

DNA damage is measured by quantification of γH2AX foci. While accepted in the field, some Pt-DNA lesions may be undetectable by this method and affect our conclusions about the extent and nature of DNA damage. In addition, ATM/ATR activity is perturbed with small molecule inhibitors, which may have off-target effects, therefore some of the observed responses may arise from modulation of other proteins.

STAR METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ralph Kleiner(rkleiner@princeton.edu).

Materials availability

Unique reagents generated in this study are available from the lead contact with a completed Materials Transfer Agreement.

Data and code availability

• All data reported in this paper will be shared by the lead contact upon request.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell lines and culture conditions

HeLa (human cervical carcinoma cell line, female), A549 (human lung carcinoma, male), HCT116 (human colon carcinoma, male) and U2OS (human osteosarcoma cell line, female) cells were cultured at 37 °C in a humidified atmosphere with 5% CO₂ in DMEM (Thermo Fisher, 11995073) supplemented with 10% fetal bovine serum (Bio-Techne, S12450H), 2 mM L-glutamine (Thermo Fisher, 25030–081), and penicillin-streptomycin antibiotics (Thermo Fisher, 15070–063). U2OS (human osteosarcoma cell line, female) cells expressing the “NBS1ΔN” truncated protein were obtained as a kind gift from Manuel Stucki (University of Zurich)⁴¹. All drug treatments were conducted on cells grown to 70% confluency. Except where noted otherwise, cisplatin and oxaliplatin drug treatments were carried out at a final concentration of 5 μM Pt drug for 6 hours. For pulsed 5-EU incorporation to visualize nucleolar transcription, cells were incubated for 1 hr in medium containing 1 mM 5-EU. To measure transcription during drug treatment, medium containing the drug would be replaced with fresh medium containing both 1 mM 5-EU and the drug compound for the final 1 hr of the treatment. To analyze cell cycle state, cells were incubated for 1 hr in medium containing 10 μM 5-EdU at the end of the corresponding drug treatment. For experiments involving co-treatment with a kinase inhibitor, cells were pre-treated with inhibitor alone for 4 hours prior to addition of fresh media containing both compounds. Unless stated otherwise, the following compounds were used at the indicated final concentrations: ATR inhibitor VE-821 (4 μM), ATM inhibitor KU-55933 (10 μM), CHK1 inhibitor GDC-0575 (300 nM), CHK1/CHK2 inhibitor AZD-7762 (300 nM), and CHK2 inhibitor BML-277 (10 μM).

METHOD DETAILS

Plasmids and siRNA Transfection

The construct for NPM1-mCherry expression was generated by PCR amplification from FM5-NPM1-mCherry (gift from Clifford Brangwynne, Princeton University) and ligation into a modified pcDNA5/FRT/TO (Life Technologies, V6520–20) vector containing an N-terminal 3xFLAG tag. HA-NLS-I-PpoI cDNA was obtained from Addgene (#46963). For transient transfection, HA-NLSI-PpoI cDNA was cloned into pcDNA5/FRT/TO vector (Life Technologies, V6520–20). DNA plasmids were transfected into cells with Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Immunofluorescence analysis of I-Ppo1 was performed 8 hr after the transfection. For silencing of protein expression by RNAi, the control siRNA (siCtrl) and siRNA against TOPBP1 (siTOPBP1) and Treacle (siTCOF) were obtained from Dharmacon (Horizon Discovery). The sequences of siRNAs used in this study are listed in Key Resources Table. Cells were grown in six-well plates 24 hr prior to transfection. The siRNA transfection was performed with 20 nM siRNA using Lipofectamine RNAiMAX (Invitrogen). Cells were split on coverslips 24 hours after the transfection and used for immunofluorescence 72 hours after transfection.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
NPM1	Thermo Fisher (Invitrogen)	Cat# 32-5200; Clone# FC-61991
Anti-DNA-RNA hybrid	Millipore Sigma	Cat# MABE1095; Clone# S9.6
TCOF1	Millipore Sigma	Cat#: HPA038237
p53	Millipore Sigma	Cat# OP43; Clone# DO-1
β-Actin	Cell Signaling	Cat# 3700; Clone# 8H10D10
UBF	Santa Cruz	Cat# sc-13125; Clone# F-9
RPA194	Santa Cruz	Cat# sc-48385; Clone# C-1
TAF I p110	Santa Cruz	Cat# sc-374551; Clone# C-10
phospho-H2AX (S193)	Millipore Sigma	Cat# 05-636; Clone# JBW301
NBS1	Novus Biologicals	Cat# NB100-143
TOPBP1	Thermo Fisher (Bethyl Laboratories)	Cat# A300-111A-M
MRE11	GeneTex	Cat# GTX70212; Clone# 12D7
phospho-RPA (S4, S8)	Thermo Fisher (Bethyl Laboratories)	Cat# A300-245A
FLAG	Millipore Sigma	Cat# F1804; Clone# M2
HA	Cell Signaling	Cat# 2367; Clone# 600
CHK1	Santa Cruz	Cat# sc-8408; Clone# G-4
phospho-CHK1 (S345)	Cell Signaling	Cat# 2348; Clone# 133D3
CHK2	Cell Signaling	Cat# 2662
phospho-CHK2 (T68)	Cell Signaling	Cat# 2197; Clone# C13C1
Cyclin A	Santa Cruz	Cat# sc-271682; Clone# B-8
Chemicals, peptides, and recombinant proteins
5-ethynyl uridine (5-EU)	Carbosynth	Cat# NE31862
2’-azidocytidine (2’-AzCyd)	Carbosynth	Cat# NA05412
THPTA	Click Chemistry Tools	Cat# 1010-100
Cy3-alkyne	Click Chemistry Tools	Cat# TA117-5
Cy3-DBCO	Click Chemistry Tools	Cat# A140-5
CX-5461	Cayman Chemical	Cat# 18392
BMH-21	Cayman Chemical	Cat# 22282
doxorubicin	Cayman Chemical	Cat# 15007
9-hydroxy-ellipticine	Cayman Chemical	Cat# 17921
GDC-0575	Cayman Chemical	Cat# 34300
AZD7762	Cayman Chemical	Cat# 11491
BML-277	Cayman Chemical	Cat# 17552
VE-821	Sigma-Aldrich	Cat# SML1415
KU-55933	Sigma-Aldrich	Cat# SML1109
Cisplatin	Sigma-Aldrich	Cat# 1134357-100MG
Actinomycin D	Sigma-Aldrich	Cat# A1410
Oxaliplatin	Sigma-Aldrich	Cat# O9512
Experimental models: Cell lines
HeLa (human, cervical carcinoma) gender: female	Kind gift from Thomas Shenk	N/A
U2OS (human, bone osteosarcoma) gender: female	Kind gift from Tom Muir	N/A
U2OS NBS1ΔN (human, bone osteosarcoma) gender: female	Kind gift from Manuel Stucki	N/A
A549 (human, lung carcinoma) gender: male	Kind gift from Alexander Ploss	N/A
HCT116 (human, colon carcinoma) gender: male	Kind gift from Tom Muir	N/A
Oligonucleotides
5’-UGGUUUACAUGUCGACUAA	Dharmacon	siCtrl
5’-ACAAAUACAUGGCUGGUUA	Dharmacon	siTOPBP1
5’-CCACCAUGGGUUGGAACUAAAUU	Dharmacon	siTCOF1
5’-Biotin-CTCCCGCGTGTGTCCTGGGGTTGACCAGAG	IDT	prHu3-Promoter Forward
5’-GGCACGGTGGCCCTCGCCGCCTTCC	IDT	prHu3-Promoter Reverse
5’-IRDye800-GCTCTTCCTATCATTGTGAAGCAGA	IDT	28S-rRNA-RT (nt 4455-4479)
5-IRDye800-TCTGGTCCGTCTTGCGCCGGTCCAAGAATTT	IDT	18S-rRNA-RT (nt 961-992)
5’-IRDye800-AAAGCCTACAGCACCCGGTAT	IDT	5S-rRNA-RT (nt 101-121)
Plasmids
pcDNA5-NPM1	This Paper	N/A
pcDNA5-HA-NLS-I-PpoI	This Paper	N/A
Software and algorithms
Prism 9 Version 9.3.1	GraphPad Software Inc.,	N/A
ImageJ	Schneider et al.	https://imagej.nih.gov/ij/

Immunofluorescence

Cells to be imaged were grown on 12 mm coverslips (Fisher Scientific, 12-545-81) in 24-well plates as described above. After treatment, cells were washed once with ice-cold PBS, fixed for 20 minutes at RT with PBS containing 3% paraformaldehyde adjusted to pH 7.3, and then washed 3 times for 5 minutes each with 1×PBS. Next, cells were permeabilized with 0.5% Triton-X in PBS for 20 minutes at RT. For click chemistry labeling of cellular RNA in 5-EU and 5-EdU treated samples, coverslips were incubated in 100 μL drops of freshly prepared CuAAC reaction mixture containing Cy3-azide (10 μM, Click Chemistry Tools, TA117–1), CuSO₄ (1 mM), THPTA ligand (2 mM), and sodium ascorbate (10 mM) in PBS. The reaction was allowed to proceed for 2 hours at room temperature in the dark. Cells were washed three times in PBS for 5 minutes each to remove free Cy3 dye. For effective visualization of nucleolar Cy3 signal for 5-EU treated cellular RNA, further immunostaining or incubation in blocking buffer was necessary to wash out nucleoplasmic signal. For immunostaining, the coverslips were blocked with 5% goat serum or 3% BSA in PBS for 1 hour, then incubated with primary antibody for 2 hours at room temperature. After washing three times with PBS for 5 minutes each, secondary antibody incubations were performed for 1 hour at RT in the dark. All antibodies used in this study are listed in Key Resources Table. The coverslips were then washed twice more with PBS for 5 minutes each, stained with Hoescht 33342 (Thermo Scientific, 1 ug/mL) for 10 minutes, washed with PBS once more, and mounted on glass microscopy slides in ProLong AntiFade Reagent (Life Technologies) and sealed with nail polish. Primary antibodies were used in the following dilutions: NPM1, 1:300; S9.6, 1:1000; TCOF1, 1:150; UBF, 1:250; RPA194, 1:100; phospho-H2AX, 1:400; NBS1, 1:150; TOPBP1, 1:200; MRE11, 1:100; phosphor-RPA, 1:100; HA, 1:1000; Cyclin A, 1:500.

In-gel fluorescence

For gel analysis of ongoing rRNA transcription, U2OS WT cells were pre-treated with indicated drug compound for 2 hr, then co-treated with drug and 1 mM 2’-azidocytidine for an additional 6 hr. Cells were harvested by scraping in 1×PBS and total RNA was extracted using TRIzol LS (Invitrogen) according to the manufacturer’s instructions. RNA samples were then reacted with Cy3 fluorophore using SPAAC conditions: 15 μg of total RNA was combined with 100 μM Cy3-DBCO (Click Chemistry Tools, A140–1) and 1X PBS in a 20 μL reaction volume, containing 0.5 μL of murine RNase inhibitor (NEB, M0314), and incubated for 2 hr at 37 °C. All reactions were purified using Zymo RNA Clean and Concentrator-5 spin columns according to the manufacturer’s instructions and separated by native gel electrophoresis using 1% TAE-agarose. Labeled RNA was visualized by in-gel fluorescence on a Typhoon FLA 9500 Fluorescent Image Analyzer Scanner (GE Healthcare) using a Cy3 filter set. Total RNA was visualized by staining with ethidium bromide.

rDNA Promoter Pulldown of Nucleolar Proteins

The biotinylated rDNA promoter (−193 to +238) was synthesized by PCR amplification using the plasmid prHu3⁵⁶ (gift from Robert Tjian, UC Berkeley) as a template. The primers used are listed in Key Resources Tables; the sense primer was biotinylated at the 5’ end. The resulting 5’-end-biotinylated DNA fragments were gel purified and immobilized on streptavidin-coated paramagnetic beads (M280 Dynabeads, Invitrogen) according to the manufacturer’s instructions. In a typical reaction⁵⁷, 750 ng of DNA was immobilized on 15 uL of beads, blocked with 5% BSA, then incubated with gentle agitation in U2OS cell nuclear extract for 30 minutes at 4 °C in 50 mM KCl – TM10i buffer (50 mM Tris HCl pH 7.4, 12.5 mM MgCl₂, 1 mM EDTA, 10% glycerol, 1 mM DTT, 0.03% NP-40) containing EDTA-free protease inhibitor cocktail (Roche). After separation using a magnetic stand, beads were washed three times with 2 reaction volumes of TM10i-0.05M KCl buffer. Captured proteins were eluted by boiling in SDS loading buffer and analyzed by immunoblotting.

Primer Extension Analysis

For primer extension analysis of Pt adducts on cellular RNA, HeLa cells were treated with the indicated Pt compounds for 12 hr, then harvested by scraping into ice-cold 1×PBS solution. Total RNA was extracted using TRIzol LS (Invitrogen) according to the manufacturer’s instructions. After DNase I treatment (NEB, M0303) and purification using Zymo RNA Clean and Concentrator-5 columns, 4 μg of total RNA was annealed to 10 pmol of IR-labeled primer (sequences listed in Key Resources Table) complementary to various segments of rRNA by heating to 65 °C for 3 min and slow cooling (0.5 °C/s) to 4 °C. Primer extension was performed using reverse transcriptase SuperScript III (Invitrogen) in the manufacturer’s reaction buffer for 1 hr at 55 °C. Remaining RNA template was degraded by addition of NaOH to a final concentration of 50 mM and heating at 80 °C for 10 min, followed by quenching with HCl. The final samples were diluted in formamide loading buffer and separated on 8% dPAGE gels containing 7 M urea. Bands were visualized using the LI-COR Odyssey Imaging System.

Western Blotting

Cells were harvested by scraping into ice-cold PBS solution, and lysed in RIPA buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 0.1% Na-deoxycholate, 0.1% SDS, protease inhibitor tablet (Roche), 1 mM PMSF). Proteins were separated in 8–12% SDS-PAGE gels and transferred to nitrocellulose membranes. Membranes were then blocked with 5% BSA in TBST (5 mM Tris-HCl, pH 7.5, 15 mM NaCl, 1% Tween-20) before application of the indicated antibodies (Key Resources Table). After incubation with IR-labeled secondary antibodies and washes in 1×TSBT, blots were imaged on a LI-COR Odyssey Imaging System. Primary antibodies were used in the following dilutions: p53, 1:1000; β-actin, 1:10,000; UBF, 1:1000; RPA194, 1:500; TAF I p110, 1:500; CHK1, 1:1000; phospho-CHK1, 1:500; CHK2, 1:1000; phospho-CHK2, 1:2000.

Cell Viability Assays

Cells were plated in 96-well culture plates (5,000 cells in 200 μL of medium per well) 16 hr prior to treatment with various concentrations of the indicated drug compounds. Cell viability after 72 hr was measured using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (MTS) (Promega) according to the manufacturer’s instructions. Absorbance was measured at 490 nm using a Synergy H1 Microplate Reader (BioTek), and cell viability was calculated as the ratio of absorbance of treated cells to that of untreated cells.

QUANTIFICATION AND STATISTICAL ANALYSIS

Graphs were generated with Prism 9 for Windows (GraphPad Software Inc., Version 9.3.1). Unless otherwise indicated, horizontal bars represent the mean values and error bars the standard deviation. In box plot graphs, boxes represent the 25–75 percentile range, and whiskers represent the 9–95 percentile range. Data points outside this range are shown individually. Statistical tests were applied as described in the Figure Legends and were calculated using Prism 9.

SIGNIFICANCE.

Pt-based drugs are an important class of anti-cancer therapeutics in widespread clinical use, but a number of key questions remain relating to their mechanism of action. Historically, Pt compounds have been proposed to function as DNA damaging agents, but recent studies have demonstrated that oxaliplatin, one of the three FDA-approved Pt drugs, is better characterized as a transcription/translation inhibitor. These findings suggest that oxaliplatin kills cells in a fundamentally different manner from other Pt drugs, however the underlying molecular mechanism remains unknown. In this work, we characterize a novel pathway that links DNA damage signaling to transcriptional inhibition in the nucleolus by oxaliplatin. We show that selective silencing of rRNA transcription in this context is mediated by the DNA damage response kinases ATM and ATR. Further, we find that oxaliplatin does not selectively target nucleolar proteins or nucleic acids, suggesting that ATM/ATR activation is induced by oxaliplatin lesions outside of the nucleolus. We draw parallels to an in-trans nucleolar DNA damage response pathway, but show that the oxaliplatin response is distinct. In this manner, our study illuminates the mechanistic diversity of nucleolar DDR signaling. Moreover, we compare oxaliplatin against other small molecule inhibitors of rRNA synthesis and find that oxaliplatin is unique in its reliance upon DNA damage signaling. We anticipate that this mechanistic discovery will expand the repertoire of biological pathways that can be targeted by further development of therapeutic inhibitory agents.

Highlights.

• Oxaliplatin specifically inhibits RNA Pol I and disrupts nucleolar morphology

• Oxaliplatin induces n-DDR factor recruitment in the absence of nucleolar DNA damage

• Inhibition of nucleolar transcription by oxaliplatin involves ATM and ATR signaling

• NBS1 is not required for nucleolar RNA Pol I inhibition by oxaliplatin

INCLUSION AND DIVERSITY

We support inclusive, diverse, and equitable conduct of research.