Pharmacological Inhibition of PARP6 Triggers Multipolar Spindle Formation and Elicits Therapeutic Effects in Breast Cancer Zebin Wang1, Shaun E

Published OnlineFirst October 8, 2018; DOI: 10.1158/0008-5472.CAN-18-1362

Cancer Translational Science Research

Pharmacological Inhibition of PARP6 Triggers Multipolar Spindle Formation and Elicits Therapeutic Effects in Breast Cancer Zebin Wang1, Shaun E. Grosskurth1, Tony Cheung1, Philip Petteruti1, Jingwen Zhang1, Xin Wang1, Wenxian Wang1, Farzin Gharahdaghi1, Jiaquan Wu1, Nancy Su1, Ryan T. Howard2, Michele Mayo1, Dan Widzowski1, David A. Scott1, Jeffrey W. Johannes1, Michelle L. Lamb1, Deborah Lawson1, Jonathan R. Dry1, Paul D. Lyne1, Edward W. Tate2, Michael Zinda1, Keith Mikule1, Stephen E. Fawell1, Corinne Reimer1, and Huawei Chen1

Abstract

PARP proteins represent a class of post-translational mod- subset of breast cancer cells in vitro and antitumor effects in vivo. ification enzymes with diverse cellular functions. Targeting In addition, Chk1 was identified as a specific substrate of PARPs has proven to be efficacious clinically, but exploration PARP6 and was further confirmed by enzymatic assays and of the therapeutic potential of PARP inhibition has been by mass spectrometry. Furthermore, when modification of limited to targeting poly(ADP-ribose) generating PARP, Chk1 was inhibited with AZ0108 in breast cancer cells, we including PARP1/2/3 and tankyrases. The cancer-related func- observed marked upregulation of p-S345 Chk1 accompanied tions of mono(ADP-ribose) generating PARP, including by defects in mitotic signaling. Together, these results establish PARP6, remain largely uncharacterized. Here, we report a proof-of-concept antitumor efficacy through PARP6 inhibi- novel therapeutic strategy targeting PARP6 using the first tion and highlight a novel function of PARP6 in maintaining reported PARP6 inhibitors. By screening a collection of PARP centrosome integrity via direct ADP-ribosylation of Chk1 and compounds for their ability to induce mitotic defects, we modulation of its activity. uncovered a robust correlation between PARP6 inhibition and induction of multipolar spindle (MPS) formation, which was Significance: These findings describe a new inhibitor of phenocopied by PARP6 knockdown. Treatment with AZ0108, PARP6 and identify a novel function of PARP6 in regulating a PARP6 inhibitor with a favorable pharmacokinetic profile, activation of Chk1 in breast cancer cells. Cancer Res; 78(23); potently induced the MPS phenotype, leading to apoptosis in a 6691–702. 2018 AACR.

Introduction spontaneous tumorigenesis, although the typical consequence of multipolar mitosis is often detrimental to dividing cells (5). The presence of supernumerary centrosomes is a common To ensure successful mitosis, aneuploid cancer cells can evolve feature of solid malignant tumors, differentiating them from several mechanisms to avoid chromosome segregation errors (6). normal cells. Centrosomes normally function as the main micro- One major mechanism involves clustering and assembling mul- tubule-organizing centers (MTOC) and play an essential role tiple centrosomes into pseudo-bipolar spindles through a tightly during cellular mitosis (1). Evidence suggests that centrosome regulated process (6, 7). Disruption of this process results in abnormalities contribute to a high degree of aneuploidy in cancer multipolar mitosis and has been shown to be a therapeutic cells and are associated with advanced tumor grade (2–4). A recent strategy to specifically kill cancer cells with amplified centrosomes report indicates centrosome amplification is sufficient to promote while sparing normal cells (8, 9). The tubulin-stabilizing agent griseofulvin and its derivative GF-15, as well as the HSET/KIFC1 inhibitor AZ82, have been reported to induce formation of multipolar spindles (MPS) to selectively kill cancer cells 1 Oncology, IMED Biotech Unit, AstraZeneca R&D Boston, Waltham, Massachu- (10–12). Through unbiased siRNA screens, members of the setts. 2Institute of Chemical Biology, Department of Chemistry, Imperial College therapeutically tractable PARPs enzyme family have also been London, London, United Kingdom. implicated in centrosome clustering and bipolar spindle forma- Note: Supplementary data for this article are available at Cancer Research tion, providing the preliminary evidence to further investigate Online (http://cancerres.aacrjournals.org/). the pharmacological inhibition of PARPs to specifically perturb Z. Wang and S.E. Grosskurth contributed equally to this article and share first mitosis in cancer cells (8, 13). authorship. PARPs are a family of 17 enzymes that catalyze the transfer of þ Corresponding Author: Huawei Chen, AstraZeneca R&D Boston, 35 Gatehouse the ADP-ribose from NAD to post-translationally modify accep- Drive, Waltham, MA, 02451. Phone: 781-839-4417; Fax. 781-839-4200; E-Mail: tor proteins. Depending on their catalytic activity, PARPs can be [email protected] further divided into poly(ADP-ribose) generating PARPs such as doi: 10.1158/0008-5472.CAN-18-1362 PARPs 1-5, catalytically inactive PARPs such as PARPs 9 and 13, 2018 American Association for Cancer Research. and mono(ADP-ribose) generating PARPs such as PARPs 6-8,

www.aacrjournals.org 6691

Wang et al.

10-12, and 14-15 (14, 15). The post-translational modification alone positive control—negative control). IC50 values were cal- of substrate proteins by PARPs has been demonstrated, for culated by fitting the percentage inhibition to a four-parameter example, by PARP1-3 to regulate numerous signaling cascades, logistic regression equation with an AstraZeneca in-house calcu- including DNA damage response, chromatin remodeling, and lation engine. transcriptional regulation, as well as by PARP5a-5b for telomere maintenance, spindle assembly, vesicular movement, and degra- Immunofluorescent multipolar spindle assay dation of the beta-catenin destruction complex (16, 17). Given Indicated cell lines were plated in 96-well plates at 7,000 cells that PARPs have functional roles in mitosis and that PARP1/2 per well and incubated at 37C overnight. The cells were treated inhibition is a successful therapeutic approach to treat homolo- with compounds ranging from 0 to 11 mmol/L for 48 hours. gous recombination defective tumors in the clinic (18, 19), The cells were fixed with 4% formaldehyde at room temperature this prompted us to perform a cell-based MPS phenotypic for 10 minutes followed with ice-cold methanol fixation for screen with compounds that have structural similarity to PARP another 10 minutes. After four washes in PBS, the cells were inhibitor scaffolds. incubated in blocking buffer for 1 hour at room temperature to Here, we report the characterization of novel small molecule reduce non-specific binding. The cells were labeled with primary PARP inhibitors that induce MPS formation in cancer cells and antibodies, 1:2,000 dilution of anti-cyclin B antibody (Thermo provide evidence that this phenotype is due to PARP6 inhibition. Fisher) and 1:4,000 dilution of anti-pericentrin antibody PARP6 is a mono (ADP-ribose) generating PARP with little (Abcam), for 16 hours at 4C. After washing with PBS four times, biological characterization. Pharmacological inhibition of PARP6 the cells were labeled with secondary antibodies, 1:200 Alexa in breast cancer cells using AZ0108 (20), an optimized inhibitor Fluor 488 anti-rabbit antibody and Alexa Fluor 594 anti-mouse developed from phenotypic screen hits, resulted in MPS forma- antibody, for 1 hour at room temperature. After washing with PBS tion, impaired cell growth, and induction of apoptosis in vitro and twice, the cells were stained with Hoechst dye for 10 minutes at in vivo. Using a high-density protein array–based ADP-ribosyla- room temperature. The cells were washed twice with PBS and then tion assay, we identified Chk1 kinase within a subset of proteins images were acquired by ImageXpress Micro (Molecular Devices) involved in regulating centrosome function that were enriched as or Operatte (PerkinElmer) High Content Screening System. direct PARP6 substrates. We confirmed PARP6 ADP-ribosylates Cyclin B label was used for scoring the mitotic cells and pericen- Chk1 directly and demonstrated Chk1 S345 phosphorylation was trin was used for scoring the centrosome number in each mitotic significantly upregulated upon PARP6 inhibition, which was cell. The percentage of increase of mitotic cells with greater than accompanied by de-activation of other mitotic proteins during 2 centrosomes compared with the DMSO control was used to the G2–M transition. Taken together, our studies demonstrate a calculate EC50 values. For CHK1 and pericentrin staining, CHK1 critical role for PARP6 in ensuring the integrity of mitosis and antibody (1:100, Cell Signaling Technology) and pericentrin provide preclinical proof-of-concept for inhibiting PARP6 as a (Abcam) antibodies were used to stain HCC1806 cells after novel cancer therapeutic strategy. AZ0108 treatment following the same protocol. Images were taken with Operetta (PerkinElmer) High Content Screening Sys- tem, and CHK1 intensity was analyzed with Harmony analysis Materials and Methods software (PerkinElmer). Chemicals and cell lines AZD2281/olaparib, AZ9482, and AZ0108 were synthesized Breast cancer cell line proliferation assays by AstraZeneca (20–22) and diluted in dimethyl sulfoxide Cell lines were sourced as previously described (23, 24) and (Sigma-Aldrich). Majority of cell lines were purchased from were cultured in RPMI-1640 medium supplemented with 10% theATCCorsomewereobtainedfromDSMZandcultured FBS (GIBCO), 2 mmol/L L-glutamine, 100 U/mL penicillin and according to the providers' instructions (detailed information 0.1 mg/mL streptomycin at 37 Cin5%CO2. Cells were treated in Supplementary Table S1). Cell lines were generally main- with compounds diluted in DMSO. Cell proliferation was deter- tained at low passages before use. mined by two methods, MTS and Sytox Green assay. Briefly, cells were seeded in 96-well plates (at a density to allow for logarithmic PARP protein enzymatic assays growth during the 72-hour assay) and incubated overnight at All assays were performed following the BPS PARP assay kit 37 C, 5% CO2. Cells were then exposed to concentrations of PARP protocols. In brief, PARP enzymatic reactions were conducted in inhibitors ranging from 0.003 to 30 mmol/L for 72 hours. For the duplicate at room temperature for 1 hour in a 96-well plate coated MTS endpoint, cell proliferation was measured by the CellTiter with histone substrate, except for TNKS1 and 2 (PARP5a and AQueous Non-Radioactive Cell Proliferation Assay (Promega) PARP5b), where GST-TNKS1 or 2, instead of histone, was coated reagent in accordance with the manufacturer's protocol. Absor- on the glutathione plate for auto–ADP-ribosylation. Reaction bance was measured with a Tecan Ultra instrument. For the Sytox þ buffer (50 mL, Tris-HCl, pH 8.0) containing NAD , biotinylated Green endpoint, Sytox Green nucleic acid dye (Invitrogen) diluted þ NAD , activated DNA, a PARP enzyme (Supplementary Table S2 in TBS-EDTA buffer was added to cells (final concentration of 0.13 þ for specific enzyme amounts and NAD concentrations) and the mmol/L) and the number of dead cells was detected using an test compound were added to start the reaction. Fifty mLof Acumen Explorer. Cells were then permeabilized by the addition streptavidin-horseradish peroxidase was added to each well and of saponin (0.03% final concentration, diluted in TBS-EDTA the plate was incubated at room temperature for 30 minutes. After buffer), incubated overnight, and a total cell count was measured. adding 100 mL of developer reagents, luminescence was measured Predose measurements were made for both MTS and Sytox Green using a BioTek Synergy 2 microplate reader. Percent inhibition endpoints, and the concentration needed to reduce 50% of the were calculated for each compound concentration with the fol- growth of treated cells relative to the untreated cells (GI50) were lowing equation: (reaction signal—negative control)/(enzyme determined using absorbance readings (MTS) or live cell counts

6692 Cancer Res; 78(23) December 1, 2018 Cancer Research

Targeting PARP6 in Breast Cancer

(Sytox Green). All cell line panel MTS results are summarized in perform functional enrichment analysis on the 230 unique Supplementary Table S3. PARP6 substrates (28, 29). Functional clusters with enrichment scores >1.3 were considered significantly over-represented for In vivo efficacy study PARP6 ADP-ribosylation. All ProtoArray results are provided in Female C.B.-17 SCID (severe combined immunodeficient) Supplementary Table S4. mice were purchased from Charles River Laboratories. Mice were housed under pathogen-free conditions in individual ventilated In vitro ADP-ribosylation assay cages (IVC) at our Association for the Assessment and Accredita- Recombinant GST-PARP6 protein was produced in sf21 cells tion of Laboratory Animal Care–accredited facility in Waltham, and purified. For the Chk1 in vitro ADP-ribosylation assay, MA. Animal studies were approved by the internal IACUC (Insti- 70 nmol/L GST-PARP6, 250 nmol/L GST-Chk1 (BPS Bioscience, tutional Animal Care and Use Committee) and reported follow- cat# 400439) or 100 nmol/L BSA (Bio-Rad, cat# 5000207), and þ ing the ARRIVE (Animal Research: Reporting In Vivo experiments) 10 mmol/L biotin-labeled NAD (BPS Bioscience, cat# 80610) guidelines (25). were diluted in 250 mmol/L Tris-HCl (pH 8.0), 15 mmol/L MDA-MB-468 (10 10^6) or HCC1806 (2 10^6) tumor MgCl2, 5 mmol/L DTT, and 0.05% Triton X-100. Reactions were cells in serum-free medium with Matrigel (1:1 ratio) were injected incubated at 30C for 3 hours, terminated by adding SDS sample subcutaneously in the right flank of 5- to 6-week-old mice in a buffer and separated by Bis-Tris gel (Life Technologies). The biotin volume of 0.1 mL. Tumor volumes (measured by caliper), animal signal was detected by using horseradish peroxidase–conjugated body weight, and tumor condition were recorded twice weekly for streptavidin (CST, cat# 3999) and visualized using PIERCE the duration of the study. The tumor volume was calculated using SuperSignal West Duro Extend Duration Substrate (cat# 34076). the formula: length (mm) width (mm)2/0.52. Mice were randomized into control and treatment groups based on tumor volumes using stratified sampling. When the mean tumor size Results reached 225 mm3, treatment was initiated via oral gavage with PARP6 suppression is responsible for inducing MPS formation either vehicle (15% captisol) or AZ0108 10 mg/kg daily for 4 and centrosome defects weeks. Growth inhibition was assessed by comparison of the The recent clinical success of the PARP inhibitors olaparib differences in tumor volume between control and treated groups. (AZD2281), niraparib, and rucaparib highlights the utility of Data were log transformed to remove any size dependency before targeting the PARP enzyme family for treating human malignan- statistical evaluation. Statistical significance was evaluated using a cies (30–32). To further explore the therapeutic potential of one-tailed, two-sample t test. inhibiting the PARP family of proteins, we performed a phenotypic cell-based compound screen using a subset of AstraZeneca's ProtoArray screen for ADP-ribosylation substrates collection of PARP inhibitors, predominantly from the phthala- ProtoArray version 5.0 (Life Technologies Thermo Fisher) is a zinone series. We chose to evaluate mitotic defects since the human protein microarray with 9,000þ full-length human pro- biological function of certain PARPs had been implicated in teins printed in tandem duplicates plus controls. ADP-ribosyla- mitosis (33, 34). The goal of the screen was to identify compounds tion experiments were performed by Life Technologies as follows. that could induce the formation of MPS, a cytotoxic phenotype First, arrays were blocked in BPS BioSci buffer (#80602 BPS for cancer cells (6, 11). We established a quantitative immuno- Biosciences) and BSA. Then, the arrays were incubated in buffer, fluorescent assay in HeLa cells to measure MPS induction based NAD/biotinylated-NAD mixture, 250 nmol/L PARP1 or 125 to on the percent increase in mitotic cells with increased centrosome 250 nmol/L PARP6 enzyme (#80501 and #80506 BPS Bios- number ( 4 centrosomes per cell), and an EC50 value was ciences), 0.8 mmol/L AZ0108, and activated DNA. Arrays were determined for all compounds in this subset. One of the initial washed; ADP-ribosylation from biotinylated-NAD was identified hits, compound AZ9482, demonstrated strong MPS induction with Alexa Fluor 647-conjugated streptavidin. Data acquisition activity with excellent potency (Fig. 1A and B). Subsequent was performed with a GenePix Pro 6.0 by Life Technologies and optimization led to AZ0108 with improved physical properties mapping alignment was performed with Alexa Anti Mouse Fluor and selectivity (20). AZ0108 potently induced the MPS phe- – Antibodies (8 positive control spots in tandem duplicates notype with an EC50 value of 46 nmol/L (Fig. 1C E). In per block). Negative control spots include enzyme buffer and contrast, PARP1/2 inhibitor AZD2281 was largely inactive in BSA (12 tandem duplicate spots per block). the same assay even at a high concentration of 3.7 mmol/L (Fig. Statistics were performed in R (http://www.R-project.org) with 1E). This result suggested that inhibition of other PARPs additional visualizations generated in Spotfire (TIBCO Software). besides PARP1/2 may be contributing to the MPS phenotype. First, Z-scores were calculated with the raw signals as in other To identify the accountable PARP(s), selectivity profiling for ProtoArray ADP-ribosylation experiments (26, 27). Note the high AZ0108 was performed in all available PARP enzymatic assays Z-score values in the negative arrays were from PCCA (propionyl and the data demonstrated that in addition to having PARP1/2 Coenzyme A carboxylase, alpha), a protein that has a biotin activity, AZ0108 also potently inhibited PARP6 (Supplemen- carboxylation domain and a biotinyl-binding domain. To iden- taryFig.S1;ref.20).Furthermore,weobservedastrong tify potential differences between negative arrays and various correlation (R2 ¼ 0.76) between MPS induction potency PARP array experiments, Mann–Whitney U tests were performed and PARP6 inhibitory activity for all compounds from the and P values were calculated. In addition, logistic regression phthalazinone series tested (Fig. 1F). In contrast, no such analysis was performed between the various individual PARP correlation was observed for other PARP enzymes we exam- and negative arrays to identify spots that correlated to a specific ined, including PARP1-3, TNKS1, and TNKS2 (Supplementary PARP enzyme. For DAVID Functional Annotation Clustering, Fig. S2). These observations suggest PARP6 inhibition may the on-line tools (http://david.abcc.ncifcrf.gov/) were used to be accountable for MPS induction. Consistent with our

www.aacrjournals.org Cancer Res; 78(23) December 1, 2018 6693

Wang et al.

ABE 150 O DMSO AZD2281 3.7 mmol/L NH 100 N O N AZ9482 cells (%) 50 N N

AZ9482 N

Increase of MPS-positive 0 HOECHST -3 -2 -1 210 PERICENTRIN 10 μm m CYCLIN B Concentration (log10, mol/L) C D * 150 O NH 100 N * O F N N * * AZ0108 F N

cells (%) 50 N F AZ0108 F

Increase of MPS-positive 0

m Concentration (log10, mol/L) F G H siGAPDH PARP6 MPS Induction scatter plot * 10.00 30 mol/L) m 4.00 r2 = 0.76

1.00 * 20 0.40

0.10 10 μm 0.04 10 siPARP6 * 0.01 0.00 0

Cellular MPS induction potency ( inductionCellular potency MPS 10 siGAPDH siPARP6

0.01 0.1 1 Percentage of mitotic cells with MPS

BPS PARP6 IC50 (mmol/L) *

HOECHST PERICENTRIN CYCLIN B

Figure 1. Identification of small-molecule PARP6 inhibitors capable of inducing MPS formation in HeLa cells. A, Dose response of AZ9482 in promoting MPS-positive cells. Mitotic cells with at least four centrosomes were scored as MPS positive. B, Chemical structure of AZ9482. C, Dose response of AZ0108 in promoting MPS-positive cells. D, Chemical structure of AZ0108. E, Representative images of HeLa cells treated with DMSO, 3.7 mmol/L AZD2281, 0.04 mmol/L AZ0108, and 0.4 mmol/L AZ0108. Cyclin B stains for mitotic cells (red) and pericentrin stains for centrosomes (green). , representative cell in enlarged box for each respective 2 panel. F, Scatter plot correlating MPS induction potency (EC50) versus PARP6 enzyme inhibition IC50 for compounds (Pearson r ¼ 0.76). Highlighted compounds are: dark blue, AZ0108; light blue, AZ9482; red, AZD2281. G, Representative immunofluorescent images taken 72 hours after transfection of siRNAs targeting GAPDH and PARP6, demonstrating PARP6 knockdown triggers MPS formation. Hoechst, DNA marker (blue); pericentrin, centrosome marker (green); and cyclin B, mitosis marker (red). , representative cell in enlarged box for each respective panel. H, Quantification of percentage of mitotic cells with MPS phenotype (>2 centrosomes per nuclei) in G, where double asterisks () represent the raw P value of <0.001.

expectation, PARP6 knockdown induced MPS formation, PARP6 inhibition impairs proliferation and induces apoptosis closely phenocopying the inhibitory effect of PARP6 inhibitors in breast cancer cell lines (Fig.1GandH).Takentogether,ourdataestablishtheMPS To evaluate the antitumor potential of the PARP6 inhibitor phenotype induced by these PARP compounds is through AZ0108, growth inhibition was examined in a panel of breast PARP6 inhibition and suggests a role for PARP6 in maintain- cancer cell lines since centrosome abnormalities have been pre- ing centrosome integrity. viously described in malignant breast tissues (35). Designating

6694 Cancer Res; 78(23) December 1, 2018 Cancer Research

Targeting PARP6 in Breast Cancer

GI50 < 1 mmol/L as an arbitrary cutoff value for defining sensitivity, the same treatment regimen gave lesser activity in HCC1806 the waterfall plot for AZ0108 clearly indicates that it is not pan- xenograft tumors (32% TGI, P < 0.05; Fig. 3A). It is noted that antiproliferative across the entire panel, as a number of breast HCC1806 is an aggressive tumor with a fast tumor doubling time. cancer lines are insensitive to treatment (Fig. 2A, top). The Treatment with AZ0108 was well tolerated with little body weight sensitivity profile of AZ0108 differs drastically from that of loss when compared with vehicle control animals (Fig. 3A). PARP1/2 inhibitor AZD2281 that is inactive for all 18 lines that To investigate whether antitumor activity of AZ0108 was asso- we have data on (Fig. 2A, bottom). This observation suggests a ciated with the induction of mitotic defects in vivo, samples of unique dependency for PARP6 in select breast cancer cell lines. MDA-MB-468 and HCC1806 xenografts tumors were collected Perturbations causing centrosome abnormalities have been pre- 48 hours after a single dose of 10 mg/kg AZ0108 and examined by viously shown to induce apoptosis in cancer cells (11) and we g-tubulin immunohistochemistry for mitosis effects. Treated speculated that PARP6 inhibition would induce apoptosis to tumor samples exhibited an increase in number of cells showing drive antitumor activity. To further evaluate the phenotypic mitotic defects, including MPS, disorganized spindles, and consequence of PARP6 inhibition, the induction of apoptosis by mitotic failure (Fig. 3B). The percent change of aberrant mitotic AZ0108 in two sensitive cell lines, HCC1806 and MDA-MB-468, nuclei with AZ0108 treatment compared with vehicle control was was followed longitudinally by monitoring the increase of caspase rather remarkable, increasing from 8% to 32% for MDA-MB-468 signal. AZ0108 induced strong apoptosis in both cell lines tumors and from 23% to 63% for HCC1806 tumors (Fig. 3C). after approximately 2 days (Fig. 2B). The PARP6 dependency in AZ0108 sensitive breast cancer cell lines was further examined Identification of Chk1 as a physiological substrate of PARP6 using PARP6 shRNA and siRNA. Consistent with the effect of To gain molecular insights into MPS induction by PARP6 AZ0108 treatment, PARP6 knockdown impaired the cell viability inhibition, we sought to identify direct substrates of PARP6. of HCC1806 and MDA-MB-468 cancer lines (Fig. 2C). It is worth Currently, there is no known PARP6 substrate reported. In addi- noting that, similar to what we have observed previously in HeLa tion, very few approaches are effective in identifying substrates of cells, PARP6 suppression either by AZ0108 treatment or by siRNA MAR-generating PARPs such as PARP6 due to the lack of a suitable mediated knockdown efficiently induced MPS formation in MAR-specific antibody. Hence, we decided to use high density HCC1806 cells (Fig. 2D and E). When additional breast cancer protein microarrays (ProtoArray) in conjunction with biotiny- þ cell lines were examined, marked induction of MPS was observed lated NAD as a donor substrate for streptavidin-based detection in BT-549 and MDA-MB-157 breast cancer cells, which showed of novel ADP-ribosylation substrates, bypassing the requirement moderate to high sensitivity to AZ0108 (Fig. 2A, Supplementary of a MAR-specific antibody (27). Protein microarrays spotted Fig. S3). In contrast, minimal increase of mitotic cells with MPS with 9,000þ full-length human proteins were incubated with was seen in MCF7 cells following AZ0108 treatment, consistent full-length PARP6 and the ADP-ribosylated proteins were iden- with insensitivity of this cell line to AZ0108 (Supplementary tified with Alexa Fluor 647–conjugated streptavidin (Z-score Fig. S3). The MPS phenotype was also not observed with difference >2.5 between the intensities of PARP6-treated arrays PARP1/2 inhibitor AZD2281 or PARP1 knockdown (Supplemen- and control, Supplementary Table S4). DAVID functional enrich- tary Fig. S4), nor with AZ0108 in the immortalized noncancerous ment analysis (https://david.ncifcrf.gov) of the ADP-ribosylated mammary gland epithelial cell line MCF10A (Supplementary proteins shows a statistically significant over-representation for Fig. S5). Altogether, these results suggest there is potential utility kinases, cytoskeletal proteins, proteins at cell junctions, and cell- of a PARP6 inhibitor for MPS induction and apoptotic cell killing cycle proteins (Table 1, showing the first biological sub-cluster in breast cancer cells. for each significant cluster). Of particular interest, a biological In an effort to identify potential biomarkers associated sub-cluster under cytoskeletal proteins involved in regulating with AZ0108 sensitivity (or insensitivity), given that clinically centrosome function is significantly enriched, which may be relevant genetic alterations or subtypes in breast cancer did proteins of interest accountable for the MPS phenotype induced not predict response, we found that high expression and genetic by PARP6 inhibition. amplification of centrosome-related proteins KIAA1429 (CENP-A Because AZ0108 also inhibits PARP1/2, a PARP1 ProtoArray centromere complex binding protein; ref. 36) and SCYL1/TEIF experiment was performed in parallel to identify PARP1 sub- (centrosomal linking protein; ref. 37) appeared to be linked strates. A similar number of proteins, 194 and 204, were found to to insensitivity (Supplementary Table S5). Although the associa- be PARylated by PARP1 or ADP-ribosylated by PARP6, respec- tions are circumstantial on their own, intriguingly we identified tively. However, there is little overlap of the putative substrates SCYL1 as a substrate for PARP6 from ProtoArray studies (see identified in this study between PARP6 and PARP1, or with below; Supplementary Table S4). Further work is needed to substrates that have been reported for other PARP proteins establish if these are bona fide insensitivity biomarkers and to including PARP2, PARP10, and PARP14 using the same Proto- define the molecular mechanism of how their elevated expression Array technology (Fig. 4A; Supplementary Table S4; refs. 26, 27). levels provide protection from PARP6 inhibition. Among the enriched centrosomal proteins, Chk1 was priori- tized for subsequent analysis since it was uniquely identified PARP6 inhibitor AZ0108 induces centrosome defects and through the PARP6 ProtoArray and has also been shown to demonstrates antitumor activity in breast cancer xenograft prevent premature activation of the cyclin-B-Cdk1 complex in models initiating mitosis (23), in addition to its role in DNA damage To evaluate the antitumor activity in vivo, we explored the effects induced centrosome amplification (38). The level of Chk1 ADP- of dosing AZ0108 in MDA-MB-468 and HCC1806 breast cancer ribosylation was diminished when the amount of PARP6 protein xenograft models. Significant antitumor activity [91% TGI (tumor used in the ProtoArray experiment was reduced, and was nearly growth inhibition), P < 0.01] was observed in MDA-MB-468 abolished when AZ0108 was included, suggesting a specific model with daily oral dosing of 10 mg/kg of AZ0108, whereas PARP6-medidated ADP-ribosylation of Chk1 (Fig. 4B, top). In

www.aacrjournals.org Cancer Res; 78(23) December 1, 2018 6695

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2018 American Association for Cancer Research. 6696 Downloaded from acrRs 82)Dcme ,2018 1, December 78(23) Res; Cancer ottaseto)o Z18o Z28 ramn 4 or) ylnBsann rd niae ioi ouain eieti gen a sdt used was 20 (green) bar, Pericentrin Scale population. (blue). mitotic DNA indicates stain (red) to staining used B was Cyclin dye hours). Hoechst (48 and treatment centrosome AZD2281 or AZ0108 or post-transfection) nestv (GI insensitive ramn cosapnlo 2bes acrcl ie.Cl ie r aeoie ssniie(GI sensitive as categorized are lines Cell lines. cell cancer breast 22 of panel a across treatment age al. et Wang AP niio Z18ipispoieain nue ppoi,adpooe P nbes acrcells. cancer breast in MPS promotes and apoptosis, induces proliferation, impairs AZ0108 inhibitor PARP6 2. Figure yhg-hogpttm-as mgn ytmIcct Zoom. Incucyte system imaging time-lapse high-throughput by available). not data (NA, iN rsRA otm elvaiiywseautda asps-rnfcinfrbes acrcl ie olwn niae eeperturbation gene indicated following lines cell cancer breast for post-transfection signi days Statistical 6 SE. at with evaluated mean was as viability presented cell Bottom, shRNA. or siRNA e uli in nuclei) per Student A C D t PARP6 mRNA 3 Day assay et hr natrs ( asterisk an where test, GI , m Viability (%) Knockdown (%) AZD2281 50 mol/L 100 100 60 60 20 20 40 80 40 80 0.03 0.06 0.33 0.66 0.01 0 0 D 0.1 3.3 6.6 50 10 33 hr obeatrss( asterisks double where , 1 GAPDH siRNA MDA-MB-468 HCC1806 5

m ZR-75-1 o/)bsdo h GI the on based mol/L) cancerres.aacrjournals.org

MDA-MB-157 AZ0108 Sensitivityprofileinbreastcancerlines PARP6 siRNA * Am05 B, HCC1806 Published OnlineFirstOctober8,2018;DOI:10.1158/0008-5472.CAN-18-1362 Con * MDA-MB-468 ﬂ ec n aps-/ ciainoe iei h rsneo MO 0no/ Z18 r3 mlLA98 captured AZ9482 nmol/L 30 or AZ0108, nmol/L 30 DMSO, of presence the in time over activation caspase-3/7 and uence PARP6 siRNA * SK-BR-3 DH05 ersnsraw represents ) SUM52PE * T-47D Control shRNA BT-20 ersn raw represent ) ﬁ ac oprn h en ftecnrladthe and control the of means the comparing cance 50 MDA-MB-453

aus otm Z28 estvt sasse ntesm sa omtars h ratcne elln panel line cell cancer breast the across format assay same the in assessed is sensitivity AZD2281 Bottom, values. MDA-MB-436 shRNA PARP6 * 1053

P BT-549

< HCC1187 o,PR6mN ncdw ef knockdown mRNA PARP6 Top, 0.05. Marginal >1 Sensitive ≤1mmol/L NA Insensitive ≥5mmol/L shRNA PARP6

* MDA-MB-231 944

D, HCC1395 * < 0.001.

ersnaieiae fHC86clsfloigPR6mN ncdw 7 hours (72 knockdown mRNA PARP6 following cells HCC1806 of images Representative HCC1419 m mmol/Land<5

m. HCC1569

E, HCC1954 0.2 μmol/L 0.5 μmol/L

Quanti JIMT-1 MCF7 mmol/L ﬁ B ainfrpretg fclswt P hntp ( phenotype MPS with cells of percentage for cation

MDA-MB-468 HCC1806 E PARP6 50 ﬁ Cell confluence (%) Cell confluence (%) 100 120 inywseautda 4husps-rnfcino indicated of post-transfection hours 24 at evaluated was ciency 20 40 80 60 10 20 30 40 50 60 70 0 0 1 0 0 Time elapsed(h) m tree ncdw a eemndb two-way a by determined was knockdown -targeted o/) agnl(1 marginal mol/L), Percentage ofmitotic cells with MPS 25 25 A, Growth 15 10 20 0 5 o,GI Top, 07 100 75 50 siCONT 100 75 50

siP

50 ARP6 aefl ltfr3dyAZ0108 3-day for plot waterfall AZ0108 0.4 mmol/L DMSO m

mol/L Caspase-positive Caspase-positive AZ0108 1.2 mmol/L cell area (%) cell area (%) 10 15 20 25 30 10 15 20 25 30 35 0 5 0 5 0 0 <

m DMSO 30 nmol/LAZ9482 30 nmol/LAZ0108 DMSO 30 nmol/LAZ9482 30 nmol/LAZ0108 AZD2281 1.2 Time elapsed(h) GI Apoptosis 25 25 50 < acrResearch Cancer 07 100 75 50 07 100 75 50 5 >

m mol/L centrosomes 2 o/) and mol/L), aawere Data . label o Published OnlineFirst October 8, 2018; DOI: 10.1158/0008-5472.CAN-18-1362

Targeting PARP6 in Breast Cancer

A B MDA-MB- 468 Tumor volume Normal mitosis Aberrant mitosis 500 Metaphase Metaphase Disorganized Multipolar Vehicle Anaphase Mitotic failure (top-view) (side-view) spindle spindle AZ0108 - 10 mpk 400

300 mean ± SEM 15 μm

Tumor volume (mm³) Tumor 200 * 10 15 2520 30 4035 45 Days after implant

MDA-MB- 468 Body weight 10 Vehicle AZ0108 - 10 mpk 5 MDA-MB-468

40 μm 0

-5 mean ± SEM (%) Body weight change -10 10 15 2520 30 4035 45 Days after implant

HCC1806 Tumor volume HCC1806 1,500 Vehicle 1,200 AZ0108 - 10 mpk

900 Vehicle 48 h AZ0108 - 10 mpk 48 h & 600 mean ± SEM 300 Tumor volume (mm³) C MDA-MB-468 HCC1806 81318 23 28 70 70 Days after implant * 60 60

HCC1806 Body weight 50 50 10 Vehicle 40 * 40 AZ0108 - 10 mpk 5 30 30 mean ± SD (%) 0 mean ± SD (%) 20 20 Aberrant mitotic nuclei Aberrant mitotic nuclei 10 10 -5 * mean ± SEM (%) 0 0

Body weight change AZ0108 AZ0108 -10 Vehicle Vehicle 8 13182328 48 h 10 mpk 6 h 10 mpk Days after implant 48 h 48 h

Figure 3. AZ0108 is efficacious in inhibiting tumor growth and causes mitotic defects in vivo. A, MDA-MB-468 and HCC1806 xenografts were established, randomized, and then dosed daily (orally) with 10 mg/kg AZ0108. Tumor volume was measured and graphed as SEM. The percentage of body weight change is relative to animal weight at the start of dosing. Statistical significance comparing the means of the vehicle and the AZ0108-treated animals was determined by a two-way Student t test, where an ampersand (&) or asterisk () represents raw P values of <0.05 or <0.01, respectively. B, PD samples were taken 48 hours after treatment. g-Tubulin immunohistochemistry staining was performed to indicate mitosis effects of AZ0108 in MDA-MB-468 and HCC1806 tumors. Representative normal and aberrant mitotic MDA-MB-468 cells are shown in the top. White-filled arrows, normal mitotic cells; gray-filled arrows, intermediate cells; black-filled arrows, aberrant mitotic cells. C, Quantification for aberrant mitotic nuclei by g-tubulin staining from tumor PD sample at respective time points (in hours) after single-dose treatment of AZ0108 in MDA-MB-468 and HCC1806 tumors. The mean and SD are plotted. Statistical significance between vehicle and AZ0108 was determined by a test of equal proportions, where an asterisk () represents raw P values of <0.01.

comparison, SFRS1, a previously reported PARP1 substrate, was Again, the modification of Chk1 was reduced when AZ0108 was PARylated by PARP1 but not by PARP6, suggesting ADP-ribosyla- included in the enzymatic assay. tion in this ProtoArray experiment is specific (Fig. 4B, bottom). To determine whether Chk1 is ADP-ribosylated by PARP6 The PARP6 mediated ADP-ribosylation of Chk1 was further intracellularly, macrodomain mAf1521 pulldown followed by confirmed by an in vitro enzymatic assay where PARP6 effectively mass spectrometry analysis was performed on cell lysates pre- used GST-Chk1 as an acceptor substrate in a ADP-ribosylation pared from MDA-MB-468 cells under different conditions, that is, þ reaction with biotinylated NAD as the donor substrate (Fig. 4C). DMSO control, nocodazole treatment (microtubule inhibitor to

www.aacrjournals.org Cancer Res; 78(23) December 1, 2018 6697

Wang et al.

Table 1. Summary of the top 15 biological functional clusters from DAVID functional enrichment for in vitro PARP6 ADP-ribosylated proteins Enrichment Cluster score DAVID functional term ADP-ribosylated proteins P Benjamini 1 15.33 Protein kinase AAK1, ABL1, ABL2, ADRBK1, AKT3, BMX, BTK, CAMK2A, 7.1E25 4.3E22 CDK7, CHEK1, CSNK1G1, CSNK1G2, DAPK3, EPHA1, EPHA2, EPHA3, EPHB1, FES, GRK6, MAP2K6, MAP4K5, MAPK1, MATK, NUAK1, PAK1, PIM1, PRKCA, PRKCI, PRKG2, PTK2, RIOK1, RPS6KB1, RPS6KB2, SCYL1, SRPK1, STK25, STK33, STK40, TEC, TTBK2 2 5.41 Pleckstrin homology-type ACAP1, ADRBK1, AKT3, ANLN, ARHGEF1, BMX, BTK, DNM2, 1.9E07 8.4E06 HOMER1, HOMER2, HOMER3, NUP50, PLCG2, RALGPS1, RALGPS2, SH3BP2, TEC 3 5.13 Tyrosine-protein kinase ABL1, ABL2, BMX, BTK, EPHA1, EPHA2, EPHA3, EPHB1, FES, 2.2E09 5.5E08 MAP2K6, MATK, PTK2, TEC 4 4.23 Cytoskeleton ABL1, ABL2, ABLIM1, ADD2, ANLN, BAG1, BMX, CDC42EP3, 1E07 2.5E05 CEP57, CHEK1, CKAP2, CSPP1, DNALI1, DNM2, DYNC1I1, EPB49, FSD1, GSN, HIP1R, HOMER1, HOMER2, HOMER3, LMNA, MAPK1, MAPRE1, MYOT, NME2, ODF2, PDLIM5, PTK2, SCYL1, SEPT9, TBCC, TMOD2, TPPP, TTBK2, TWF1, TWF2, USP2 Centrosome BMX, CEP57, CHEK1, CSPP1, FSD1, MAPRE1, NME2, ODF2, 0.00029 0.0103 SCYL1, USP2 5 3.82 Cytoskeletal protein binding ABLIM1, ADD2, ANLN, ANXA2, BAIAP2, CDC42EP3, DNM2, 1E06 2.5E05 DYNC1I1, EPB49, FMNL1, GSN, HIP1R, HOMER2, MAPRE1, MYOT, NME2, PDLIM5, TMOD2, TPPP, TWF1, TWF2 6 3.53 AGC-Kinase C-terminal ADRBK1, AKT3, GRK6, PRKCA, PRKCI, PRKG2, RPS6KB1, 1.8E06 0.00018 RPS6KB2 7 2.90 Non-membrane spanning protein ABL1, ABL2, BMX, BTK, FES, MATK, PTK2, TEC 3.4E07 9E06 tyrosine kinase activity 8 2.49 Protein kinase cascade BTK, C7ORF16, DAPK3, MAP2K6, MAP4K5, MAPK1, NFKBIA, 0.00017 0.0422 PAK1, PDE6H, PIK3R1, PRKCA, RPS6KB1, RPS6KB2, SRPK1, TEC 9 1.90 Phosphoprotein phosphatase inhibitor PPP1R8, PPP1R2P9, C7ORF16, PPP1R10 0.00379 0.0735 10 1.79 Regulation of protein polymerization ADD2, ARF6, CDKN1B, EPB49, GSN, MAPRE1, TPPP 0.00017 0.0364 11 1.59 Cell junction CAMK2A, DNM2, HOMER1, HOMER2, HOMER3, LIN7B, 0.00353 0.046 MPP7, NPHP1, PAK1, PTK2, RPS6KB1, STXBP5 12 1.57 D111/G-Patch AGGF1, GPKOW, RBM17 0.0233 0.351 13 1.56 Phosphoinositide binding ANXA2, PICALM, PIK3R1, HIP1R, NCF1, TULP3 0.00553 0.0945 14 1.55 ErbB signaling pathway ABL1, ABL2, AKT3, CAMK2A, CDKN1B, MAPK1, PAK1, 3.2E09 3.1E07 PIK3R1, PLCG2, PRKCA, PTK2, RPS6KB1, RPS6KB2 15 1.51 Cell cycle ANLN, CDCA3, CDK7, CDKN1B, CHAF1B, CHEK1, CKAP2, 0.00127 0.0187 FSD1, MAPK1, MAPRE1, NCAPG, PIM1, SEPT9 NOTE: Highlighted in blue is a cytoskeletal subcluster for centrosomal proteins.

induce prometaphase arrest), and cotreatment of nocodazole In summary, the results from ProtoArray screening, biochemical with AZ0108. Because no antibody is currently available to confirmation using isolated proteins, and cellular investigation effectively recognize MAR post-translational modification, the through mass spectrometry analyses provide solid evidence that macrodomain mAf1521 was used as a tool to pulldown mono- Chk1 is a physiological substrate of PARP6. and poly-ADP-ribose unit(s) containing proteins (39). Gel elec- trophoresis coupled with silver staining of the macrodomain- PARP6 inhibition by AZ0108 results in elevated p-S345 Chk1 pulldown material identified a clear difference in staining inten- and reduced mitotic signaling in vitro and in vivo sity within the 50-75 kDa region (band highlighted in Fig. 4D, Previous reports noted that at interphase, Chk1 is able to top) between the different treatment conditions, showing prevent premature activation of the cyclin-B–Cdk1 complex, and increased staining with nocodazole treatment (Fig. 4D, compare overexpression of centrosome-targeted Chk1 prohibited centro- Lane 2 vs. 1) that was reversed with AZ0108 cotreatment (Fig. 4D, some separation as well as induced polyploidization (23, 40). compare Lane 3 vs. 2). Mass spectrometry analyses were per- Given the importance of Chk1 function in regulating mitosis and formed on the highlighted band excised from the gel and a our finding that Chk1 is a substrate of PARP6, we set out to handful of protein candidates including Chk1 were identified investigate the effects of PARP6 inhibition on the G2–M phase (Fig. 4D, bottom). Consistent with our expectation, Chk1 was transition in HCC1806 cells following sequestration at prome- only present in the band under the treatment condition taphase using nocodazole. Activation of mitosis-related proteins with nocodazole alone. Because the samples were enriched for such as FoxM1, Aurora kinases B and C, and Histone H3 were ADP-ribosylated proteins with macrodomain pulldown, these observed when cells were entering mitosis (the 0 hour time results indicated that Chk1 is ADP-ribosylated when cell cycle point for phosphorylated proteins, and in the case of FoxM1, the was perturbed to induce prometaphase arrest and this ADP- slower migrating band corresponding to the phosphorylated, ribosylation was prevented with AZ0108 treatment, suggesting active form), all of which are known to be phosphorylated (i.e. that Chk1 ADP-ribosylation in cells is likely mediated by PARP6. activated) during the G2–M transition (Fig. 5A, under DMSO).

6698 Cancer Res; 78(23) December 1, 2018 Cancer Research

Targeting PARP6 in Breast Cancer

A C Streptavidin Coomassie blue detection PARP6 ++ ++ ++++ Figure 4. Biotin-NAD+ - +++ - +++ Identification of Chk1 as a specific BSA - + -- - + -- PARP6 -- -- substrate of PARP6. A, Venn diagram GST-Chk1 ++ ++ 204 proteins AZ0108 -- -+ ---+ summary of proteins ADP-ribosylated PARP1 250 KD by PARP1, PARP2, PARP6, PARP10, 194 proteins 150 KD and PARP14 highlighting the low 100 KD PARP6 PARP6 proportion of proteins ADP- 75 KD GST-Chk1 GST-Chk1 ribosylated by multiple PARPs. B, PARP10 PARP14 12 proteins Z-score bar charts with individual 58 proteins 50 KD protein spot scatter plots for Chk1 and PARP2 37 KD SFRS1 across different ProtoArray 46 proteins experimental conditions. Values for 25 KD the ADP-ribosylation of Chk1 by 20 KD PARP6 and SFRS1 by PARP1 are 15 KD statistically different from the no PARP enzyme condition with raw P B D IP: Macrodomain values <0.05 () using a Mann– mAf1521 ProtoArray protein Z-score value Nocodazole - + Whitney U Test. C, Coomassie blue + 4 AZ0108 - - + staining and Western blot analysis of PARP6 activity tests using 3 þ * 250 KD biotinylated NAD as donor, GST-Chk1 150 KD as substrate, with and without 2 100 KD NPM1 AZ0108. ADP-ribosylation was 1 75 KD ANXA2 detected with the streptavidin- Chk1 Chk1 conjugated system. D, Top, 0 50 KD DSC1 MYO5B silver staining of macrodomain- 3 37 KD pulldown materials prepared from MDA-MB-468 cells treated with DMSO 2 control, nocodazole for mitotic * Lane 1 Lane 2 Lane 3 Full protein name MW [kDa] synchronization, or nocodazole plus SFRS1 1 -NPM1-Nucleophosmin, isoform 2 29.4 - ANXA2 - Annexin A2 38.6 0.3 mmol/L AZ0108. Bottom, - Chk1 - S/T-protein kinase Chk1 54.4 identification of candidate proteins -DSC1-Desmocolin-1, isoform 1B 93.8 0 - MYO5B - Myosin 5B 82.4 from the highlighted band from mass spectrometry analysis. No PARP PARP1 PARP6 PARP6 PARP6 control 250 nmol/L 250 nmol/L 125 nmol/L 125 and 800 nmol/L AZ0108

Interestingly, when AZ0108 was cotreated with nocodazole intensity of Chk1 within the cells did not change (Supplementary marked activation of Chk1 (upregulated p-S345 Chk1 level) and Fig. S6). These observations provide further evidence that deactivation of FoxM1, Aurora kinases, and Histone H3, were ADP-ribosylation mediated by PARP6 affects the activation and observed (Fig. 5A, under AZ0108). However, a similar effect was localization of Chk1 in cancer cells. not observed with PARP1/2 inhibitor AZD2281 (Fig. 5A, under To extend the impact of Chk1 phosphorylation findings in vivo, AZD2281). Taken together, these data strongly suggest that Chk1 tumors were examined for pharmacodynamic effects on nuclear activation is regulated by PARP6 via ADP-ribosylation during p-S345 Chk1 by immunohistochemistry following a single oral mitosis and is accompanied by inhibition of activated mitotic administration of 10 mg/kg AZ0108 to female C.B.-17 SCID mice signaling. bearing HCC1806 sub-cutaneous xenografts. Staining for nuclear To examine the impact of PARP6 inhibition on Chk1 activation p-S345 Chk1 was significantly higher for AZ0108-treated tumors under endogenous settings without cell-cycle perturbation, relative to vehicle controls by immunohistochemistry (Fig. 5C). MDA-MB-468 and HCC1806 breast cancer cells were treated with When quantified, an approximately six-fold increase of nuclear AZ0108 and levels of phosphorylated and total Chk1 were eval- p-Chk1 staining was observed in the treatment group at 6 hours uated (Fig. 5B). AZ0108 treatment resulted in dose-dependent (Fig. 5D). The increase of p-Chk1 staining was still evident at 24 h increase of p-S345 Chk1 without altering the total protein expres- but largely dissipated when the drug was cleared from the system sion in both cell lines. Profound induction was achieved with (48 hours). Taken together, learnings from these molecular 300 nmol/L AZ0108 at 24 hours with no further increase observed investigations provide a mechanistic description of PARP6 when incubation time was extended to 48 hours. Consistent with regulating Chk1 activation in mitosis and may account for our previous observations in HCC1806 cells under nocodazole the MPS phenotype associated with PARP6 inhibition from perturbation conditions (Fig. 5A), p-S10 histone H3 was reduced AZ0108 treatment. with AZ0108 treatment in a concentration-dependent manner in both cell lines (Fig. 5B). Considering the role of Chk1 in G2–M checkpoint and centrosome maintenance, we further examined if Discussion AZ0108 treatment leads to accumulation of Chk1 to the centro- We uncovered PARP6 as a molecular target for a subset of somes. Costaining of Chk1 with centrosome marker pericentin PARP compounds identified through a phenotypic screen for revealed that increasing amount of Chk1 is localized to the MPS formation in cancer cells. Initial evidence came from the centrosome following AZ0108 treatment, whereas the overall observation of a robust correlation between potency of PARP6

www.aacrjournals.org Cancer Res; 78(23) December 1, 2018 6699

Wang et al.

A B 24 h 48 h MDA-MB-468 HCC1806 MDA-MB-468 HCC1806 Hours after DMSO AZ0108 AZD2281 nocodazole release 033003300330 Chk1 Chk1 p-S345 p-S345 Total Chk1 Chk1 Histone H3 Histone H3 p-S10 p-S10 FoxM1 Histone H3 Figure 5. AurA p-T288 PARP6 inhibition by AZ0108 results AurB p-T232 Vinculin AurC p-T198 in elevated p-S345 Chk1 and reduced in vitro in vivo GAPDH mitotic signaling and . 1.00 Chk1 0.88 0.93 0.85 0.57 A, Western blot analysis of mitotic 0.45 0.39 a-Tubulin p-S345/Total 0.29 0.21 0.26 0.06 0.07 proteins and their activation status in HCC1806 cells that were mitotically synchronized with 40 ng/mL Histone H3 0.92 1.00 0.99 0.57 0.66 0.68 nocodazole for 16 hours and released p-S10/Total 0.39 0.44 0.33 0.17 0.19 0.23 for 0, 3, or 30 hours while treated with DMSO, 0.3 mmol/L AZ0108, AZ0108 or 0.3 mmol/L AZD2281. B, Representative Western blots of 0 nmol/L 0 nmol/L 30 nmol/L 0 nmol/L 0 nmol/L 300 nmol/L 30 nmol/L 30 nmol/L 30 nmol/L 300 nmol/L 300 nmol/L 300 nmol/L p-S345 and total Chk1 as well as p-S10 C 6 h 6 h 24 h 48 h and total histone H3 for MDA-MB-468 and HCC1806 cells following 24 or 48 hours of AZ0108 treatment at 0, 30, and 300 nmol/L (top). Quantification of Western blot intensity of p-Chk1 and p-S10 histone H3 normalized against total protein (bottom, N ¼ 3). C, IHC staining of p-S345 Chk1 for PD HCC1806 samples from HCC1806 tumors for 6, 24, and 48 hours after single dose of 50 mm 10 mg/kg AZ0108 treatment. D, Quantification of C. The mean and SD are plotted. Statistical significance Vehicle AZ0108 10 mpk between vehicle and AZ0108 at the same time point was determined by a D HCC1806 25 test of equal proportions, where an * asterisk () represents raw P value of 20 <0.01.

15 - Positive nuclei 10

mean ± SD (%) 5

p-S345 Chk1 0 Vehicle 6 24 48 h 6 h AZ0108 - 10 mpk

enzyme inhibition and activity in MPS induction. This corre- can induce the MPS phenotype when used at a high concentration lation was associated with PARP6 but not with other PARP (35), we believe this is likely to be an off-target effect. This notion enzymes examined, including PARP1, PARP2, PARP3, TNKS1, is further corroborated by the observation that PARP1 knockdown and TNKS2. We further supported this finding with PARP6 failed to induce the MPS phenotype. Finally, similar to AZ0108, knockdown, which phenocopied the induction of MPS forma- PARP6 knockdown led to impaired viablility in cancer cells. tion in cancer cell lines. Thus, AZ0108 is a suitable small-molecule probe to investigate We further characterized AZ0108, the first potent and orally the phenotypic consequence of MPS induction in cancer available PARP6 inhibitor, identified through optimization of models, allowing further evaluation of PARP6 as a potential phenotypic screening hits. While AZ0108 also carries PARP1/2 therapeutic target. activity, several lines of evidence support that AZ0108 is still a AZ0108 has a distributive antitumor response in a panel of suitable PARP6 probe compound. Firstly, AZ0108 resides nearly breast cancer lines, with activity associated with strong induction on the regression line on the correlation plot between PARP6 of the MPS phenotype and cancer cell apoptosis. The sensitivity inhibition and MPS induction, suggesting PARP6 inhibition is a profile of AZ0108 is differentiated from that of PARP1/2 inhibitor driver for the MPS phenotype. Secondly, PARP1/2 inhibitor AZD2281, further underscoring the different phenotypic conse- AZD2281 (olaparib), which lacks PARP6 activity, does not induce quences of inhibiting PARP6 or PARP1/2. The antitumor activity MPS formation even at 3.7 mmol/L concentration, compared with of AZ0108 is translatable in vivo, with efficacy observed in two the significant MPS effect with AZ0108 at concentrations as low as breast cancer xenograft models, MDA-MB-468 and HCC1806. 0.046 mmol/L. Although there is a report that a PARP1/2 inhibitor The treated tumors exhibited a statistically significant increase in

6700 Cancer Res; 78(23) December 1, 2018 Cancer Research

Targeting PARP6 in Breast Cancer

aberrant mitotic cells. While these observations are encouraging, if ADP-ribosylation creates a structural hindrance that interferes several gaps remain. Additional work around biomarkers such with the interaction of Chk1 with its modifying kinase and as the centrosome-related proteins KIAA1429/CENP-A and phosphatase. Additional mechanistic work is warranted to SCYL1/TEIF that we found to be associated with breast cancer further delineate the molecular mechanism of how PARP6- cell line insensitivity to AZ0108 would be important to identify catalyzed Chk1 ADP-ribosylation affects its phosphorylation. potential models and patients that would most likely benefit from Collectively, our findings provide strong evidence support- PARP6 inhibitors. Additional pharmacological investigation ing a novel therapeutic strategy by inhibiting PARP6. We regarding the extent and duration of PARP6 inhibition necessary identified and characterized the first small-molecule inhibitor for optimal MPS induction and cancer cell apoptosis, as well as for PARP6, AZ0108. Both compound and genetic inhibition rational combinations to augment antitumor activity, would also of PARP6 function caused mitotic defects in breast cancer be beneficial. Toxicology studies will be required to determine the cells. PARP6 was shown to directly ADP-ribosylate proteins potential for mechanism-based toxicity in normal tissues. Finally, that function as kinases, cytoskeletal proteins, and centroso- further chemistry efforts to design an even more selective PARP6 mal proteins such as Chk1, providing a mechanistic basis inhibitor would be desirable for realizing optimal therapeutic for the mitotic defects observed in breast cancer cells with benefit while minimizing any off-target toxicity. AZ0108. Theothernotablelearningfromourstudyistheidentifica- tion of Chk1 as a PARP6 substrate, with its activation putatively Disclosure of Potential Conflicts of Interest modulated through PARP6-mediated ADP-ribosylation. Chk1 S.E. Grosskurth is a senior scientist—Bioinformatics at AstraZeneca. D.A. plays prominent roles in mediating DNA damage response Scott has ownership interest (including stock, patents, etc.) in AstraZeneca. signaling and regulating cell-cycle checkpoints during the S, J.R. Dry has ownership interest (including stock, patents, etc.) in AstraZeneca. P.D. Lyne has ownership interest (including stock, patents, G2, and M phases of the cell cycle. Chk1 activation is tightly etc.) in AstraZeneca. M. Zinda reports receiving other commercial research regulated for proper M phase entry and progression, and support and has ownership interest (including stock, patents, etc.) in inhibition of Chk1 has been shown to lead to various mitotic AstraZeneca. S.E. Fawell has ownership interest (including stock, patents, abnormalities and kinetochore defects (23, 24). Initially iden- etc.) in AstraZeneca. No potential conflicts of interest were disclosed by the tified from ProtoArray, Chk1 was further confirmed as a PARP6 other authors. substrate by a biochemical ADP-ribosylation assay using isolated recombinant PARP6 and Chk1 proteins. Additional sup- Authors' Contributions port was provided by indirect evidence from a cellular exper- Conception and design: Z. Wang, S.E. Grosskurth, J. Zhang, X. Wang, D. Widzowski, D.A. Scott, M.L. Lamb, J.R. Dry, P.D. Lyne, M. Zinda, K. Mikule, iment in which the ADP-ribosylation of Chk1, induced by S.E. Fawell, C. Reimer, H. Chen mitotic arrest, was reversed by PARP6 inhibitor AZ0108. In Development of methodology: Z. Wang, S.E. Grosskurth, P. Petteruti, J. Zhang, breast cancer models both under in vitro and in vivo conditions, X. Wang, J. Wu, N. Su, R.T. Howard, M. Mayo, M.L. Lamb, J.R. Dry, K. Mikule, AZ0108 treatment induces robust phosphorylation of S345 of C. Reimer, H. Chen Chk1, a well-established Chk1 activation biomarker associated Acquisition of data (provided animals, acquired and managed patients, with cell-cycle checkpoints and mitotic defects including cen- provided facilities, etc.): Z. Wang, P. Petteruti, J. Zhang, W. Wang, F. Gharahdaghi, J. Wu, N. Su, R.T. Howard, J.W. Johannes, D. Lawson, trosome amplification (38), and de-activation of other key K. Mikule, C. Reimer, H. Chen mitosis proteins including FoxM1, histone H3, and Aurora Analysis and interpretation of data (e.g., statistical analysis, biostatistics, kinases. These observations support a novel mechanism of computational analysis): Z. Wang, S.E. Grosskurth, P. Petteruti, J. Zhang, PARP6 regulating Chk1 activation in mitosis, and perhaps in W. Wang, F. Gharahdaghi, J. Wu, N. Su, R.T. Howard, M. Mayo, D. Widzowski, other phases of cell cycle as well, via ADP-ribosylation and thus D.A. Scott, M.L. Lamb, J.R. Dry, K. Mikule, S.E. Fawell, C. Reimer, H. Chen provide a mechanistic basis for MPS induction by PARP6 Writing, review, and/or revision of the manuscript: Z. Wang, S.E. Grosskurth, T. Cheung, P. Petteruti, J. Zhang, W. Wang, F. Gharahdaghi, D.A. Scott, inhibition. It is possible that deregulated Chk1 phosphoryla- – M.L. Lamb, P.D. Lyne, E.W. Tate, M. Zinda, K. Mikule, S.E. Fawell, C. Reimer, tion induced by PARP6 inhibition alters the Chk1 CDK1 H. Chen pathway during mitosis and therefore affects centrosome clus- Administrative, technical, or material support (i.e., reporting or organizing tering. Interestingly, Chk1 activity was reported to be involved data, constructing databases): T. Cheung, M. Mayo, D. Widzowski, C. Reimer in correcting merotelic kinetochore attachment to ensure prop- Study supervision: T. Cheung, J. Wu, D. Widzowski, J.R. Dry, E.W. Tate, er spindle checkpoint signaling by regulating Aurora-B, MCAK, M. Zinda, K. Mikule, S.E. Fawell, C. Reimer, H. Chen Kif2b, and Hee1 (41), which also suggest that Chk1 activity Acknowledgments could directly impact the centrosome clustering process. This R.T. Howard and E.W. Tate thank AstraZeneca and the Engineering and notion is consistent with the observation that the expression Physical Sciences Research Council of the UK for support through the Imperial level of KIAA1429/CENP-A, one of the key drivers of centro- College Centre for Doctoral Training (grant EP/L015498/1). We thank Jane some clustering, is a negative correlate for AZ0108 sensitivity. Cheng for in vivo technical assistance and the anonymous reviewers for insight- However, the association of Chk1 activation with centrosome ful comments. amplification would suggest another possible mechanism for PARP6-induced MPS phenotype (38). Further mechanistic The costs of publication of this article were defrayed in part by the studies are required to define the contribution of Chk1- payment of page charges. This article must therefore be hereby marked mediated defective centrosome clustering, aberrant centrosome advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate amplification, and perhaps other mechanisms, to MPS forma- this fact. tion when PARP6 is inhibited. It is unclear whether the inter- play between Chk1 phosphorylation and ADP-ribosylation Received May 2, 2018; revised August 23, 2018; accepted September 26, 2018; results from direct competition for the site of modification or published first October 8, 2018.

www.aacrjournals.org Cancer Res; 78(23) December 1, 2018 6701

Wang et al.

References 1. Bornens M. Centrosome composition and microtubule anchoring 23. Kramer A, Mailand N, Lukas C, Syljuasen RG, Wilkinson CJ, Nigg EA, et al. mechanisms. Curr Opin Cell Biol 2002;14:25–34. Centrosome-associated Chk1 prevents premature activation of cyclin- 2. Fukasawa K. Centrosome amplification, chromosome instability and B-Cdk1 kinase. Nat Cell Biol 2004;6:884–91. cancer development. Cancer Lett 2005;230:6–19. 24. Wilsker D, Petermann E, Helleday T, Bunz F. Essential function of Chk1 can 3. Fukasawa K. Oncogenes and tumour suppressors take on centrosomes. be uncoupled from DNA damage checkpoint and replication control. Proc Nat Rev Cancer 2007;7:911–24. Natl Acad Sci U S A 2008;105:20752–7. 4. D'Assoro AB, Lingle WL, Salisbury JL. Centrosome amplification and the 25. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving development of cancer. Oncogene 2002;21:6146–53. bioscience research reporting: the ARRIVE guidelines for reporting animal 5. Levine MS, Bakker B, Boeckx B, Moyett J, Lu J, Vitre B, et al. Centrosome research. J Pharmacol Pharmacother 2010;1:94–9. amplification is sufficient to promote spontaneous tumorigenesis in 26. Troiani S, Lupi R, Perego R, Depaolini SR, Thieffine S, Bosotti R, et al. mammals. Dev Cell 2017;40:313–322.e5. Identification of candidate substrates for poly(ADP-ribose) polymerase-2 6. Godinho SA, Pellman D. Causes and consequences of centrosome abnor- (PARP2) in the absence of DNA damage using high-density protein malities in cancer. Philos Trans R Soc Lond B Biol Sci 2014;369. microarrays. FEBS J 2011;278:3676–87. 7. Quintyne NJ, Reing JE, Hoffelder DR, Gollin SM, Saunders WS. Spindle 27. Feijs KL, Kleine H, Braczynski A, Forst AH, Herzog N, Verheugd P, et al. multipolarity is prevented by centrosomal clustering. Science 2005;307: ARTD10 substrate identification on protein microarrays: regulation of 127–9. GSK3beta by mono-ADP-ribosylation. Cell Commun Signal 2013;11:5. 8. Kwon M, Godinho SA, Chandhok NS, Ganem NJ, Azioune A, Thery M, et al. 28. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis Mechanisms to suppress multipolar divisions in cancer cells with extra of large gene lists using DAVID bioinformatics resources. Nat Protoc centrosomes. Genes Dev 2008;22:2189–203. 2009;4:44–57. 9. Fielding AB, Lim S, Montgomery K, Dobreva I, Dedhar S. A critical role of 29. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: integrin-linked kinase, ch-TOG and TACC3 in centrosome clustering in paths toward the comprehensive functional analysis of large gene lists. cancer cells. Oncogene 2011;30:521–34. Nucleic Acids Res 2009;37:1–13. 10. Pannu V, Rida PC, Celik B, Turaga RC, Ogden A, Cantuaria G, et al. 30. Kaufman B, Shapira-FrommerR,SchmutzlerRK,AudehMW, Centrosome-declustering drugs mediate a two-pronged attack on inter- Friedlander M, Balmana~ J, et al. Olaparib monotherapy in patients phase and mitosis in supercentrosomal cancer cells. Cell Death Dis 2014;5: with advanced cancer and a germline BRCA1/2 mutation. J Clin Oncol e1538. 2015;33:244–50. 11. Raab MS, Breitkreutz I, Anderhub S, Rønnest MH, Leber B, Larsen TO, et al. 31. Mirza MR, Monk BJ, Herrstedt J, Oza AM, Mahner S, Redondo A, et al. GF-15, a novel inhibitor of centrosomal clustering, suppresses tumor cell Niraparib maintenance therapy in platinum-sensitive, recurrent ovarian growth in vitro and in vivo. Cancer Res 2012;72:5374–85. cancer. N Engl J Med 2016;375:2154–2164. 12. Wu J, Mikule K, Wang W, Su N, Petteruti P, Gharahdaghi F, et al. Discovery 32. Swisher EM, Lin KK, Oza AM, Scott CL, Giordano H, Sun J, et al. Rucaparib and mechanistic study of a small molecule inhibitor for motor protein in relapsed, platinum-sensitive high-grade ovarian carcinoma (ARIEL2 Part KIFC1. ACS Chem Biol 2013;8:2201–8. 1): an international, multicentre, open-label, phase 2 trial. Lancet Oncol 13. Chang P, Coughlin M, Mitchison TJ. Tankyrase-1 polymerization of poly 2017;18:75–87. (ADP-ribose) is required for spindle structure and function. Nat Cell Biol 33. Chang P, Coughlin M, Mitchison TJ. Interaction between Poly(ADP- 2005;7:1133–9. ribose) and NuMA contributes to mitotic spindle pole assembly. Mol Biol 14. Gibson BA, Kraus WL. New insights into the molecular and cellular Cell 2009;20:4575–85. functions of poly(ADP-ribose) and PARPs. Nat Rev Mol Cell Biol 2012;13: 34. Perdoni F, Bottone MG, Soldani C, Veneroni P, Alpini C, Pellicciari C, et al. 411–24. Distribution of centromeric proteins and PARP-1 during mitosis and 15. Hottiger MO, Hassa PO, Luscher€ B, Schuler€ H, Koch-Nolte F. Toward a apoptosis. Ann N Y Acad Sci 2009;1171:32–7. unified nomenclature for mammalian ADP-ribosyltransferases. Trends 35. Lingle WL, Lutz WH, Ingle JN, Maihle NJ, Salisbury JL. Centrosome Biochem Sci 2010;35:208–19. hypertrophy in human breast tumors: implications for genomic stability 16. Vyas S, Chang P. New PARP targets for cancer therapy. Nat Rev Cancer and cell polarity. Proc Natl Acad Sci U S A 1998;95:2950–5. 2014;14:502–9. 36. Obuse C, Yang H, Nozaki N, Goto S, Okazaki T, Yoda K. Proteomics 17. Schreiber V, Dantzer F, Ame JC, de Murcia G. Poly(ADP-ribose): novel analysis of the centromere complex from HeLa interphase cells: functions for an old molecule. Nat Rev Mol Cell Biol 2006;7:517–28. UV-damaged DNA binding protein 1 (DDB-1) is a component of the 18. Helleday T, Bryant HE, Schultz N. Poly(ADP-ribose) polymerase (PARP-1) CEN-complex, while BMI-1 is transiently co-localized with the centromeric in homologous recombination and as a target for cancer therapy. Cell Cycle region in interphase. Genes Cells 2004;9:105–20. 2005;4:1176–8. 37. Gong Y, Sun Y, McNutt MA, Sun Q, Hou L, Liu H, et al. Localization of 19. Farmer H, McCabe N, Lord CJ, Tutt AN, Johnson DA, Richardson TB, et al. TEIF in the centrosome and its functional association with centrosome Targeting the DNA repair defect in BRCA mutant cells as a therapeutic amplification in DNA damage, telomere dysfunction and human cancers. strategy. Nature 2005;434:917–21. Oncogene 2009;28:1549–60. 20. Johannes JW, Almeida L, Daly K, Ferguson AD, Grosskurth SE, Guan H, 38. Bourke E, Dodson H, Merdes A, Cuffe L, Zachos G, Walker M, et al. DNA et al. Discovery of AZ0108, an orally bioavailable phthalazinone PARP damage induces Chk1-dependent centrosome amplification. EMBO Rep inhibitor that blocks centrosome clustering. Bioorg Med Chem Lett 2007;8:603–9. 2015;25:5743–7. 39. Jungmichel S, Rosenthal F, Altmeyer M, Lukas J, Hottiger MO, Nielsen ML. 21. Menear KA, Adcock C, Boulter R, Cockcroft XL, Copsey L, Cranston A, et al. Proteome-wide identification of poly(ADP-Ribosyl)ation targets in differ- 4-[3-(4-cyclopropanecarbonylpiperazine-1-carbonyl)-4-fluorobenzyl]- ent genotoxic stress responses. Mol Cell 2013;52:272–85. 2H-phthalazin- 1-one: a novel bioavailable inhibitor of poly(ADP- 40. Loffler H, Rebacz B, Ho AD, Lukas J, Bartek J, Kr€amer A. Chk1-dependent ribose) polymerase-1. J Med Chem 2008;51:6581–91. regulation of Cdc25B functions to coordinate mitotic events. Cell Cycle 22. Menear KA, Adcock C, Alonso FC, Blackburn K, Copsey L, Drzewiecki J, 2006;5:2543–7. et al. Novel alkoxybenzamide inhibitors of poly(ADP-ribose) polymerase. 41. Petsalaki E, Zachos G. Chk1 and Mps1 jointly regulate correction of Bioorg Med Chem Lett 2008;18:3942–5. merotelic kinetochore attachments. J Cell Sci 2013;126:1235–46.

6702 Cancer Res; 78(23) December 1, 2018 Cancer Research

Pharmacological Inhibition of PARP6 Triggers Multipolar Spindle Formation and Elicits Therapeutic Effects in Breast Cancer

Zebin Wang, Shaun E. Grosskurth, Tony Cheung, et al.

Cancer Res 2018;78:6691-6702. Published OnlineFirst October 8, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-18-1362

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2018/10/06/0008-5472.CAN-18-1362.DC1

Cited articles This article cites 40 articles, 8 of which you can access for free at: http://cancerres.aacrjournals.org/content/78/23/6691.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/78/23/6691.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/78/23/6691. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.