Combination of PARP inhibitor and temozolomide to suppress chordoma progression

Abstract

Chordoma, a malignant bone cancer, is highly resistant to conventional therapeutic approaches; this greatly limits radio- and chemotherapeutic options and disease management. In the present study, we investigated three patient-derived chordoma cell lines to elucidate the molecular mechanism of resistance to therapeutics. An in vitro high-throughput chemical screening assay and an in vivo xenograft model were used to identify novel chemosensitizers for chordoma. We found that patient-derived chordoma cell lines recapitulated disease phenotypes, which were highlighted by robust resistance to medical therapy manifested as lack of DNA damage accumulation. Mechanistically, the PARP DNA repair pathway was found to play a central role in this resistance. Chemical screening confirmed that PARP inhibitors could strikingly enhance temozolomide (TMZ) therapy in chordoma cells. Combining the FDA-approved PARP inhibitor, olaparib, with chemotherapeutics not only potentiated DNA damage accumulation, cell cycle arrest, and apoptosis in vitro but also suppressed chordoma xenograft expansion in vivo. We conclude that combining PARP inhibition with TMZ could be an effective therapeutic approach for the clinical management of chordoma.

Introduction

Chordoma, a rare bone-derived neoplasm that accounts for 1–4% of all primary malignant bone tumors, is associated with poor prognosis. It originates from notochordal remnants and occurs along the axial skeleton [1]. Chordoma has benign features like slow growth, but also aggressive characteristics such as predilection local invasion and distant metastasis. Thus, it is a difficult disease to manage clinically, particularly as tumors frequently arise near critical neurovascular structures such as the skull base and spinal cord [2].

Due to the distinct pathogenesis and disease manifestation, current therapeutic options for chordoma are limited [3]. Complete radical resection is associated with the best local control, as compared with subtotal resection and radio/chemotherapy. However, complete resection is generally difficult to achieve due to the complicated anatomical location and/or the extent of tumor spread. Radiation therapy is a common alternative when complete surgical resection is not possible. Photon, proton, and heavier charged particle carbon therapy can delay disease progression [4–6]. However, despite the advances in other therapeutic approaches, conventional chemotherapy is frustratingly ineffective for chordoma treatment, which greatly limits the management of this bone cancer. Considering that conventional chemotherapeutic agents only exert modest cytotoxic effects on chordoma cells, the identification of chemosensitizers has become an urgent need to improve treatment.

Inhibitors of DNA repair are a class of small molecular compounds that suppress the intrinsic repair mechanism. In normal cells such as somatic cells or stem cells, DNA repair machinery protects the integrity of the genome from naturally occurring genotoxic agents by removing nucleobase adducts and repairing single- and/or double-stranded DNA breaks [7]. However, in transformed cancer cells, DNA repair pathways are frequently exploited to mitigate damage induced by radiation or chemotherapies. For example, poly ADP ribose polymerase 1 (PARP) plays a key role in the base excision repair pathway (BER), which effectively removes damaged nucleobases in cancer cells [8]. Many lines of evidence have shown that PARP1 is important for resistance to therapy in cancer cells and, as such, determines the response rate and disease outcomes [9, 10]. Combining PARP inhibitors with conventional chemotherapy has been increasingly studied in translational and clinical research on several types of human malignancies [11–13].

In the present study, using patient-derived chordoma cells, we revealed a central role for the PARP DNA repair pathway in resistance to therapy in chordoma cells. Chemotherapy-induced DNA damage is effectively scavenged by the PARP/BER pathway, preventing DNA breaks and cell death. Upon combining the FDA-approved PARP inhibitor olaparib, the therapeutic effect of the conventional alkylating agent temozolomide was profoundly enhanced, as evidenced by remarkable increases in DNA damage, DNA fragmentation, and cellular apoptosis. This combination regimen also effectively suppressed chordoma xenograft growth, which was accompanied by elevated cytotoxicity, suggesting a novel therapeutic approach for the clinical management of chordoma.

Material and methods

Cell culture

Patient-derived chordoma cell lines, U-CH1, UM-Chor1, and CH22, were kindly provided by the Chordoma Foundation [14–16]. Human glioma cell line U251 MG was purchased from Sigma in 2015. Human glioma cell line U87 MG was purchased from ATCC in 2015. Cells were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin, and streptomycin (Thermo Fisher).

Immunofluorescence

Immunofluorescent staining was performed as previously described [17]. Cells were fixed in 4% PFA for 15 min and penetrated with 0.3% Triton X-100. Cells were labeled with primary antibodies followed by fluorescence-conjugated secondary antibody. Cells were visualized in a Zeiss 780 confocal microscope. The primary antibodies used in this study include γH2A.X (CST, 1:200) and BrdU (BD Biosciences, 1:200).

Apoptosis analysis

Cellular apoptosis was analyzed by annexin V/PI flow cytometry kit according to the manufacturer’s protocol (Thermo Fisher). In brief, cells were harvested and incubated with a mixture of annexin V-FITC and PI for 20 min on ice. Cells were analyzed through a FACSCanto II (BD) flow cytometer.

Cell viability test

Cell viability was tested using CCK-8 (Dojindo) per manufacturer’s protocol. Ten microliters of CCK-8 was added into cell culture media and incubated for 2–3 h. CCK-8 absorbance was measured by an Epoch plate reader (BioTek) at OD 450 nm.

Combination index

The dose-response curve of temozolomide (TMZ) or olaparib alone and their combination was measured in chordoma cell lines U-CH1 and UM-Chor1. TMZ and olaparib were combined at 100:1 ratio. The cells were treated for 48 h and cell viability was measured by CCK-8 assay. The combination index (CI) was calculated to determine the synergistic effect using COMPUSYN software (www.combosyn.com) [18].

Chemical screening assay

CH22 and U-CH1 cells were plated in triplicate in 96-well plates. One hundred and seventy-seven chemical probes (MedChem Express), which target key elements in DNA repair-related pathways, were used in the present study. For a complete list of compounds and their targets, see Fig. 2a and Supplementary table. Cells were treated with each of the probe at a concentration of 0.5 μM (monotherapy), or a combination therapy including individual probe and TMZ (100 μM). Cell viability was investigated 72 h after initial treatment, through CCK-8 assay as described above. The potentiation of chemotherapy was determined by viability (monotherapy)/viability (combination therapy). Data was plotted using the software GraphPad Prism.

Fig. 2.

PARP DNA repair plays a central role in the therapy resistance in chordoma cells. a Illustration of chemical library used for in vitro screening. b Illustration for combination effect for each chemical probe with TMZ. Data was plotted using the viability in combination therapy and monotherapy (left panel), or a ratio of viability and p value (right panel). Screening hit was designated as fold change (F.C.) > 2, p value < 0.05. C. Immunoblotting assay showed that small interference RNA reduced PARP expression, which led to depletion of pADPR and elevated γH2A.X in CH22 and U-CH1 cells. β-actin was used as loading control. d Comet assay showed that PARP1 RNA interference sensitized chordoma cells for TMZ treatment (100 μM, 48 h), evidenced by the appearance of comet tail formed by fragmented DNA. e Quantification of tail moment from comet assay. PARP suppression by small interference RNA synergized with TMZ treatment for elevated DNA damage accumulation. **p < 0.01

Western blot

Protein was extracted from cultured cell using RIPA lysis buffer supplemented with protease inhibitor cocktail (Thermo Fisher). Protein samples were resolved on NuPAGE 4–12% minigels (Thermo Fisher) and transferred to PVDF membrane (Millipore). The membrane was probed with primary antibody, and the expression of targeted protein was visualized using HRP-conjugated secondary antibody and chemiluminescence method. The primary antibodies used in this study include pADPR (Abcam, ab14459, 1:2000), PARP1 (Abcam, ab32138, 1:2000), γH2A.X (Proteintech, 10859-1-AP, 1:1000), and β-actin (Sigma, A5441, 1:2000).

DNA fragmentation test

DNA fragmentation test was performed as previously described [19]. Genomic DNA was isolated from cells using DNeasy Blood and Tissue Kit (QIAGEN) according to the manufacturer’s manual. Five hundred nanograms of DNA was resolved on 4–20% Novex TBE gel (Thermo Fisher). The gel was stained with SYBR Safe DNA dye (Thermo Fisher) and visualized using a Bio-Rad ChemiDoc Imaging System.

Alkaline comet assay

The comet assay was performed as previously described [20]. In brief, cells were collected and resuspended in molten agarose. The suspension was resolved at 0.6 V/cm for 30 min in alkaline electrophoresis buffer. Genomic DNA was then labeled with SYBR Safe solution and visualized under a Zeiss LSM 710 confocal microscope. The tail moment was measured using comet assay software and GraphPad Prism.

Cell cycle analysis

One million cells were harvested and fixed using 75% ethanol overnight at − 20 °C. Cells were treated with RNase A and the DNA content was probed with propidium iodide (Thermo Fisher). Cell cycle progression was analyzed by a FACSCanto II (BD) flow cytometer.

BrdU incorporation assay

To assess cellular proliferation, BrdU (Sigma) was directly added into growth media at a concentration of 10 μM for 2–3 h. Cells were fixed with 4% PFA for 15 min in room temperature and permeabilized with 0.3% Triton X-100. After being treated with 2-N HCl and 0.1-M sodium borate, cells were probed with BrdU antibody (BD Biosciences). The cells were then visualized under a Zeiss LSM 710 confocal microscope.

Xenograft

Eight-week-old female NSG mice (JAX Lab) were injected into the right flank with 8 million U-CH1 cells resuspended in 100-μL PBS. Once the tumors reached over 200 mm³, mice were randomly allocated into four groups and treated with (1) DMSO, (2) Ola, (3) TMZ, or (4) Ola + TMZ. The olaparib dosage was 20 mg/kg, i.p., for a total of 24 days. TMZ dosage was 25 mg/kg, p.o., 2 × 5 day cycle. Xenograft size was measured using vernier calipers. At the end of the study, the mice were sacrificed, and xenografts were harvested for weighing and histology analysis.

Immunohistochemistry

Five-micron tissue sections were prepared from a formalin-fixed, paraffin-embedded tissue specimen. The sections were deparaffinized and subjected to heat-induced antigen retrieval. The sections were then incubated with primary antibody overnight at 4 °C, followed by HRP-conjugated secondary antibody. Target proteins were visualized using a Vectastain ABC kit and DAB kit (Vector Laboratories). The primary antibodies used in the present study include Ki67 (Abcam, 1:200), PCNA (Santa Cruz, 1:200), pADPR (Abcam, 1:200), PARP1 (Abcam, 1:200), and γH2A.X (Proteintech, 1:200).

TUNEL

TUNEL assay was performed using the DeadEnd Colorimetric TUNEL System (Promega) following the manufacturer’s protocol. The slides were detected using DAB and counterstained using hematoxylin.

Statistics

Statistics between two groups was performed using Student’s t test. For analysis involving more than two groups, data was processed using the one-way ANOVA test. The results were presented as mean ± SEM. A p value < 0.05 was considered statistically significant. All analyses were performed using GraphPad Prism software (v7.01).

Results

Patient-derived chordoma cells are highly chemoresistant

To investigate the molecular basis of resistance to therapy in chordoma, we utilized three patient-derived chordoma cell lines, namely, CH22, U-CH1, and UM-Chor1. As an initial test, we assessed cellular vulnerability to temozolomide (TMZ), a DNA alkylating agent that is commonly used to treat various types of malignancies. We found that a high dose of TMZ (100 μM) resulted in minimal DNA damage in chordoma cells (Fig. 1a, b). The lack of response to TMZ treatment was confirmed through annexin V/PI apoptosis analysis. TMZ led to substantial apoptotic changes in glioma-derived cells such as U251 and U87, whereas obvious cell death was not observed in any of the chordoma-derived cell lines with the same treatment (Fig. 1c). We identified a substantial population of apoptotic cells with U251 and U87 cell lines, whereas rates of apoptosis were less than 10% in chordoma cells using the same therapy (Fig. 1d). Moreover, cell viability assays demonstrated that a lethal dose (100 or 200 μM) did not result in a loss of viability in chordoma-derived cells (Fig. 1e).

Fig. 1.

Patient-derived chordoma cells minimally respond to TMZ therapy. a Immunofluorescence staining assay showed lack of γH2A.X puncta in chordoma cell lines when receiving TMZ treatment (100 μM, 48 h). Bar = 10 μm. b Quantification of γH2A.X puncta in CH22 and UM-Chor1 cells. Cells were treated with TMZ for 48 h. c Annexin V/PI apoptosis analysis showed lack of apoptotic changes in chordoma cells when treated with TMZ (100 μM, 48 h). U251 and U87 glioma-like cells were used as positive control. d Quantification of apoptotic cells shown in c. Glioma-like cells, but not chordoma cells, showed apoptotic changes to TMZ treatment. **p < 0.01. e CCK-8 cell viability test showed that TMZ treatment (100 or 200 μM, 48 h) caused cytotoxicity in glioma-like cells, but not chordoma cells

The PARP DNA repair pathway plays a central role in chordoma cell chemoresistance

Intrinsic DNA repair pathways play key roles in maintaining the integrity of the genome, whereas they are exploited by cancer cells when confronting genotoxic challenges such as radio- and/or chemotherapy [21]. Several lines of evidence have suggested that the PARP/BER pathway can effectively remove DNA nucleobase adducts, preventing DNA damage and cytotoxicity mediated by DNA alkylating agents such as TMZ, cyclophosphamide, and carmustine [22]. In the present study, we sought to understand whether the DNA repair pathway is responsible for the remarkable chemoresistance of chordoma cells. First, we performed a chemical screening assay involving 177 chemical probes that target 12 DNA repair-related pathways (Fig. 2a). Data were plotted to highlight candidate compounds that potentiate TMZ-induced cytotoxicity in CH22 or U-CH1 chordoma cells (fold change > 2; p < 0.05). Among these compounds, PARP inhibitors such as veliparib, rucaparib, olaparib, and AZD-2461 profoundly enhanced TMZ-mediated cytotoxicity (Fig. 2b, Supplementary table). To validate this finding, we investigated the function of the PARP DNA repair pathway in chordoma cells with TMZ treatment. We found that the poly (ADP-ribose) polymer (pADPR), which indicates the activation of PARP DNA repair pathway, was robustly expressed in CH22 and U-CH1 cells. Suppressing PARP1 expression in these cell lines using small interfering RNA not only eliminated PARP1/pADPR expression but also increased TMZ-induced DNA damage (calculated by densitometric analysis of γH2A.X signal as TMZ + siPARP1 vs. TMZ; for CH22, + 41.4%; for U-CH1, + 85.9%. Figure 2c). Notably, the reduction of pADPR and elevated DNA damage were salvaged by ectopic expression of PARP1 (Supplementary Figure 1A). For further validation, we analyzed DNA fragmentation in chordoma cells using the comet assay. We showed that TMZ did not induce DNA fragmentation in all chordoma cells. However, suppressing PARP1 expression with siRNA led to a substantial increase in DNA damage in these cells, as evidenced by the emergence of a comet tail and elevated tail moment (Fig. 2d, e). The essential role of the PARP DNA repair pathway not only explains the strikingly high resistance to chemotherapy observed in chordoma cells but also suggests that PARP inhibition could represent a novel sensitizing approach for this type of cancer.

Combining PARP inhibitors with TMZ induces chordoma cell cytotoxicity

PARP inhibitors are a group of small molecular inhibitors that were developed from structural mimetics of nicotinamides. Our preliminary findings showed that the FDA-approved PARP inhibitors rucaparib and olaparib exhibited outstanding sensitizing effects when combined with TMZ in both chordoma cell lines tested. Therefore, we further evaluated the chemosensitizing effect of olaparib in chordoma cells. First, we analyzed changes in cell cycle and revealed that a combination of TMZ and olaparib caused remarkable G2/M arrest in all three chordoma cell lines, suggesting that accumulated DNA damage could result in cell cycle arrest (Fig. 3a). Additionally, comet assays showed that TMZ/olaparib led to a substantial increase in fragmented DNA in chordoma cells (Fig. 3b). The tail moment increased by 74.8-fold (CH22), 32.1-fold (U-CH1), and 39.7-fold (UM-Chor1), when compared with that of TMZ monotherapy (Fig. 3c). Elevated DNA damage was also confirmed by γH2A.X staining, which showed an extensive accumulation of DNA damage puncta in TMZ/olaparib-treated cells (Fig. 3c, e). In contrast, TMZ or olaparib monotreatment did not result in obvious DNA damage puncta. Accordingly, we confirmed that this combination could abolish pADPR/PARP expression, which was accompanied by elevated γH2A.X expression in all chordoma cells (Fig. 3f). Finally, we found that chordoma cell expansion was suppressed by combination therapy, as evidenced by BrdU incorporation (Fig. 3g).

Fig. 3.

PARP inhibitor potentiates TMZ therapy. a Cell cycle analysis showed increased G2/M arrest in chordoma cells treated with TMZ and olaparib (Ola, TMZ 100 μM, Ola 0.5 μM, 48 h) combination regimen. b Comet assay showed that TMZ/Ola combination regimen (TMZ 100 μM, Ola 0.5 μM, 48 h) caused DNA fragmentation in chordoma cells. c Quantification of tail moment shown in b. Fragmented DNA was significantly increased by combination therapy. **p < 0.01. d Immunofluorescence staining assay showed accumulated γH2A.X puncta in chordoma cells treated with TMZ/Ola combination regimen (TMZ 100 μM, Ola 0.5 μM, 48 h). Bar = 10 μm. e Quantification of γH2A.X puncta shown in d. TMZ/Ola treatment led to significant accumulation of DNA damage. **p < 0.01. f Immunoblotting assay showed that Ola (0.5 μM, 48 h) sensitized TMZ (100 μM, 48 h), evidenced by abolished pADPR formation and elevated γH2A.X expression. β-actin was used as loading control. g BrdU incorporation assay showed that TMZ/Ola treatment (TMZ 100 μM, Ola 0.5 μM, 48 h) reduced cellular proliferation in chordoma cells, as compared with TMZ alone. Bar = 20 μm

In addition to its effect on the accumulation of DNA damage and cell cycle arrest, we also discovered that olaparib sensitized chordoma cells to chemotherapeutics, resulting in cytotoxicity and apoptotic changes in chordoma cells. Phase contrast microscopy showed a remarkable loss of adherent cells when cells received prolonged TMZ/olaparib treatment (Fig. 4a). Accordingly, by performing a dose-response assay, we observed significantly enhanced TMZ cytotoxicity when this agent was combined with low-dose olaparib (Fig. 4b). For U-CH1 cells, the TMZ IC₅₀ decreased from 851.6 to 189.8 μM, and for UM-Chor1 cells, the IC₅₀ was reduced from 690.5 to 198.6 μM, when combined with 2-μM olaparib. Additionally, calculation of the combination index showed potent synergistic effect when olaparib was introduced in the treatments (for U-CH1, Fa = 0.5, CI = 0.21; for UM-Chor1, Fa = 0.5, CI = 0.4. Figure 4c, d). We also confirmed that the loss of viable cells was through TMZ-induced apoptosis, as annexin V/PI analysis showed substantial increases in the population of apoptotic cells (Fig. 4e); specifically, greater than 80% of chordoma cells proceeded to apoptosis with combination therapy (Fig. 4f).

Fig. 4.

TMZ/Ola combination therapy prompted cytotoxicity in chordoma cells. a Phase contrast microscopy showed remarkable cell loss in chordoma cells when treated with TMZ/Ola (TMZ 100 μM, Ola 0.5 μM, 48 h). Bar = 20 μm. b Dose-response curve showed synergistic effect of TMZ and Ola treatment for 48 h. c The Fa-CI plot of TMZ and Ola combination treatment in chordoma cell lines. CI, combination index; Fa, fraction affected. d The isobologram at Fa = 0.5 of TMZ and Ola combination treatment in chordoma cell lines. e Annexin V/PI apoptosis analysis showed elevated cellular apoptosis in chordoma cells receiving TMZ/Ola treatment (TMZ 100 μM, Ola 0.5 μM, 48 h). f Quantification of the apoptosis fraction shown in c. TMZ/Ola treatment significantly increased cellular apoptosis in chordoma cells. ***p < 0.001

A combination of PARP inhibition and temozolomide suppresses chordoma xenograft growth in vivo

To determine whether the PARP inhibitor could improve the efficacy of chemotherapeutics toward chordoma, we developed a xenograft model as previously described [23]. U-CH1 cells were subcutaneously injected into NSG mice. When tumor mass reached 200 mm³, animals were randomly allocated into four treatment groups and received therapy as indicated in Fig. 5a. We found that the TMZ/olaparib combination therapy significantly altered tumor growth, as evidenced by a significant reduction in tumor size and weight (Fig. 5b–e). Further, we performed histologic analysis of xenografts from each treatment group. We found that TMZ and olaparib monotherapy led to a slight decrease in the expression of proliferative markers such as Ki67 and PCNA. However, a combination regimen resulted in a marked reduction in proliferative markers, in accordance with the previously observed cell cycle arrest (Fig. 5f). Moreover, we demonstrated that pADPR expression was extensively suppressed in olaparib treatment groups. In addition, upon combining this approach with TMZ treatment, we observed an extensive increase in cytotoxic markers such as γH2A.X and TUNEL (Fig. 5g). Therefore, through this in vivo study, we concluded that TMZ/olaparib treatment successfully reduces chordoma xenograft expansion through enhanced DNA damage, cell cycle arrest, and cytotoxicity.

Fig. 5.

TMZ/Ola regimen suppress chordoma xenograft in vivo. a Schematic illustration showed therapeutic strategy in NSG mice with U-CH1 subcutaneous xenograft. b Body weight measurement of the NSG mice during the treatment period. c Tumor size measurement showed reduced U-CH1 xenograft growth in TMZ/Ola-treated group. **p < 0.01. d Tumor size measurement of U-CH1 xenograft in the 64th day after initial implantation. **p < 0.01. e Tumor weight measurement of U-CH1 xenograft. *p < 0.05, **p < 0.01. f Histology analysis on chordoma xenograft. Immunohistochemistry showed reduced expression of Ki67 and PCNA in TMZ + Ola-treated group. Bar = 100 μm. g Immunohistochemistry showed reduced expression of pADPR in TMZ + Ola-treated group. The staining of γH2A.X and TUNEL was found elevated in the combination group. Bar = 100 μm

Discussion

In the present study, we found that patient-derived tumor cells recapitulate the disease phenotype of chordoma, highlighted by resistance to therapeutics and reduced DNA damage accumulation. Mechanistically, the PARP/BER pathway plays a key role in the detoxification of chemotherapy-derived DNA adducts, preventing cells from cell cycle arrest and cell death. Combining a pharmacological PARP inhibitor with chemotherapeutics remarkably enhanced DNA damage in chordoma cells, which not only suppressed tumor cell proliferation and promoted apoptotic cell death but also suppressed chordoma xenograft growth in vivo. Our findings thus highlight a novel therapeutic approach for chordoma patients, especially when other therapeutic modalities have been exhausted.

Chordomas are bone-derived tumors that are highly resistant to chemotherapy. Radical resection is the optimal choice for therapy; however, this option is not applicable when cancers reside in complicated anatomic structures and/or exhibit invasive/metastatic characteristics. Thus, improving the efficacy of chemotherapy is key to improving disease outcomes. In the present study, we employed three patient-derived chordoma cell lines, which were found to consistently recapitulate the resistance to genotoxic therapy as observed in the clinic (Fig. 1). Mechanistically, this resistance is probably in part due to their characteristic slow growth, which allows sufficient time for the activation of intrinsic DNA repair mechanisms to detoxify DNA damage. Additionally, we noticed that the PARP/BER pathway is highly active in chordoma cells, preventing the accumulation of DNA damage and ameliorating its associated consequences (Fig. 2c–e). These findings led us to explore the possibility of combining a PARP inhibitor with conventional chemotherapy to overcome chemoresistance and improve the treatment of chordoma.

To identify possible sensitizing agents, we employed a non-biased chemical library screen based on known therapeutic agents that target different DNA repair mechanisms (Fig. 2a). For example, the ATM/ATR pathway is activated by double-strand DNA breaks, which leads to H2A.X phosphorylation and cell cycle arrest [24]. RAD51 is a highly conserved protein that assists in the repair of DNA double-strand breaks. RAD51 overexpression has been found to be related to chemoresistance in several types of human malignancies [25, 26]. In the present study, we used TMZ as the genotoxic agent and evaluated the synergistic effect with inhibitors of DNA repair enzymes. Temozolomide is an oral chemotherapy agent, which introduces DNA damage through DNA alkylation or methylation [8]. Our recent findings suggest that the combination of TMZ and PARP inhibitor achieved superior cytotoxicity in IDH1-mutated glioma and advanced neuroendocrine tumor [13, 27, 28]. Similarly, several recent publications also indicate that TMZ may be useful in aggressive cancers other than glioma [29–32]. Among the compounds screened, PARP inhibitors comprised a group of outliers that led to superior sensitization to TMZ treatment (Fig. 2b). These compounds have been demonstrated to be effective for malignancies with intrinsic DNA repair deficiency, such as BRCA1/2-mutated ovarian or breast cancers [33, 34]. The application of PARP inhibitors as chemosensitizers has not been widely applied in the clinic, presumably due to the complexities associated with combination therapies, such as identifying optimal dosage and preclinical models [8]. Combining PARP inhibitors with conventional chemotherapy was previously found to greatly improve the efficacy of the latter, which in turn was able to exert tumor suppressive effects at a relatively low dosage [27]. In the present study, we suppressed PARP activity using either small interfering RNA or an FDA-approved PARP inhibitor, olaparib. With both, we observed more severe DNA damage and breaks in chordoma cells (Fig. 2c–e; Fig. 3). This is in accordance with many previous studies showing that PARP synergizes with TMZ to achieve improved therapeutic effects. Importantly, our study showed that with TMZ/olaparib, accumulated DNA damage resulted in cytotoxicity and cellular apoptosis (Fig. 4).

To guide future clinical studies, we developed xenograft models using U-CH1 cells. Consistent with the clinical manifestation of chordoma, U-CH1 xenografts showed high resistance to conventional therapy, as evidenced by minimally altered tumor growth and the slow accumulation of DNA damage. Combining olaparib with chemotherapeutics successfully altered disease outcome, as shown by a reduction in tumor mass expansion (Fig. 5). Additionally, histologic analysis confirmed enhanced therapeutic efficacy with combination therapy. The suppression of proliferative markers such as Ki67 and PCNA was consistent with the observed xenograft growth inhibition. In contrast, increases in cytotoxic markers such as γH2A.X and TUNEL confirmed the enhanced cytotoxicity associated with TMZ/olaparib-induced tumor suppression. Importantly, the combination regimen suppressed chordoma expansion in experimental animals (Fig. 5b). This suggests that the risk of side effects might be lower than that with monotherapy, which requires high dosages. The present study thus provides proof-of-concept evidence regarding a novel therapeutic approach that involves PARP inhibitors and genotoxic therapy, which could represent a future chemotherapeutic option for chordoma.

A limitation in the present study is that high dosage of TMZ (25 mg/kg, p.o. × 5, two cycles) was used in the preclinical animal model to achieve cytotoxicity and improvement of disease outcome. One possible reason is that in the present study, chordoma cell lines are generally chemoresistant. In addition, we introduced therapeutics when the xenografts were fully established (200 mm³); therefore, the TMZ/olaparib combination may exhibit less potency compared with treatment before the formation of tumor mass. Future study is encouraged to evaluate whether pharmacologic doses of TMZ and PARP inhibitor could achieve clinical benefit in patients with chordoma. Additionally, exploring more potent genotoxic agent for combination regimen may provide better solution for chordoma chemotherapy.

Key messages.

• The PARP DNA repair pathway enhances chemoresistance in chordoma cells.

• Combining PARP inhibitors with genotoxic agents induces chordoma cell cytotoxicity.

• PARP inhibitor combining with temozolomide suppresses growth of chordoma in vivo.