Histone chaperone FACT Complex inhibitor CBL0137 interferes with DNA damage repair and enhances sensitivity of Medulloblastoma to chemotherapy and radiation

Abstract

Medulloblastoma (MB) is a malignant pediatric brain tumor with a poor prognosis. Post-surgical radiation and cisplatin-based chemotherapy have been a mainstay of treatment, which often leads to substantial neurocognitive impairments and morbidity, highlighting the need for a novel therapeutic target to enhance the sensitivity of MB tumors to cytotoxic therapies. We performed a comprehensive study using a cohort of 71 MB patients’ samples and pediatric MB cell lines and found that MB tumors have elevated levels of nucleosome remodeling FACT (FAcilitates Chromatin Transcription) complex and DNA repair enzyme AP-endonuclease1 (APE1). FACT interacts with APE1 and facilitates recruitment and acetylation of APE1 to promote repair of radiation and cisplatin-induced DNA damage. Further, levels of FACT and acetylated APE1 both are correlate strongly with MB patients’ survival. Targeting FACT complex with CBL0137 inhibits DNA repair and alters expression of a subset of genes, and significantly improves the potency of cisplatin and radiation in vitro and in MB xenograft. Notably, combination of CBL0137 and cisplatin significantly suppressed MB tumor growth in an intracranial orthotopic xenograft model. We conclude that FACT complex promotes chemo-radiation resistance in MB, and FACT inhibitor CBL0137 can be used as a chemo-radiation sensitizer to augment treatment efficacy and reduce therapy-related toxicity in high-risk pediatric patients.

1. Introduction

Medulloblastoma (MB) is an embryonal tumor of the cerebellum and the most common malignant pediatric brain tumor. MB is a heterogeneous tumor and are characterized by four major molecularly distinct subgroups: Wingless (WNT), Sonic Hedgehog (SHH), and MYC-driven Group 3 and Group 4 [1]. Further, based on intertumoral heterogeneity within subgroups, MB can be divided into 12 different subtypes [2]. Although the current treatment consisting of neurosurgical resection and high doses of chemo/radiation therapy have considerably improved the overall survival, they cause serious side effects, such as permanent neurological sequelae and disability, leading to devastating consequences on the quality of life of pediatric cancer survivors. This underscores the urgent need for novel therapies that would reduce treatment-related toxicity while maintaining high cure rate [3].

Both radiation and chemotherapeutic drug cisplatin exert their cytotoxic effects by directly inducing DNA damage in cells. Cisplatin induces intrastrand and interstrand cross-links in DNA which are primarily repaired by nucleotide excision repair (NER) pathway [4]. On the other hand, radiation induces isolated DNA lesions including Apurinic/apyrimidinic (AP) site, single-strand breaks (SSBs) and double-strand breaks (DSBs). The multifunctional DNA repair enzyme AP-endonuclease 1 (APE1) initiates the repair of AP sites and SSBs, which constitute ~65% of the DNA damage caused by radiotherapy, via the base excision repair (BER) pathway [5]. Efficient repair of DNA lesions by BER and NER pathways can promote cell survival and induce drug resistance in tumor cells. Previous reports demonstrated that APE1 overexpression and its activity is associated with poor response to radiation and chemotherapy in MB [6]. Moreover, studies have shown a role of APE1 and BER pathway in promoting cisplatin resistance [7, 8]. The key enzymes or proteins involved in repair of AP sites, SSBs, and cisplatin-induced DNA cross-link damage in naked DNA have well characterized [9, 10]. However, it is unknown what other factors are necessary to facilitate repair in the context of nucleosome/chromatin in cellulo, and whether these factors can be targeted to increase the efficacy of chemo-radiation therapy.

FACT is a nucleosome remodeling protein complex that has histone chaperone activity and facilitates transcription in chromatin [11]. The FACT complex, a heterodimer of Structure-Specific Recognition Protein1 (SSRP1) and Suppressor of Ty (SPT16), facilitates the removal and deposition of histone H2A/H2B in nucleosomes during transcription initiation and elongation [11, 12]. Increasing evidence suggests that the FACT complex plays a role at sites of damaged DNA in chromatin in cells [13, 14]. SPT16 is found to accelerate the exchange of H2A/H2B at sites of UV-induced DNA damage, and is involved in NER pathway [13]. SSRP1 is recruited to SSB in a PARP-dependent manner and retained at DNA damage sites by N-terminal interactions with the DNA repair protein XRCC1 [14]. Curaxins (CX), a class of small molecule drugs with broad anticancer activity, destabilize nucleosomes and inhibit FACT functions in chromatin [15, 16]. The second generation CX, CBL0137 which recently entered into phase I clinical trial for hematological cancer and melanoma, modulates several important pathways through inhibition of FACT activity [17, 18]. A recent study also demonstrated that FACT-targeted CBL0137 works effectively on suppressing group 3 MB growth in vivo [19].

In this study, we seek to investigate whether FACT functionally interacts with APE1 and facilitates DNA damage repair in MB, and if targeting FACT with CBL0137 sensitizes MB to cisplatin and radiation therapy. Our study demonstrates a novel role of the FACT complex in promoting chemo-radiation resistance in MB via modulating both DNA damage repair and gene expression, and shows that FACT inhibitor CBL0137 can be used to improve the efficacy of interventions for MB and outcomes in high-risk patients.

2. Materials and methods

2.1. Cell culture and treatments

Human MB cell lines DAOY, D283 and SVGp12 were purchased from ATCC (Manassas, VA, USA). HD-MB03 and ONS-76 cell lines were purchased from DSMZ, Germany and Sekisui-XenoTech (KS, USA), respectively. UW228 cells were kindly provided by Dr. John Silber from University of Washington, Seattle. All cell lines were authenticated using STR DNA profiling by ATCC and were tested for mycoplasma contamination using MycoSensor PCR Assay Kit (Santa Clara, CA, USA).

2.2. Patient tissue samples and Immunohistochemical (IHC) analysis

We obtained 11 human MB tissues from the University of Nebraska Medical Center (UNMC) Biobank and 60 MB samples from the Department of Pathology of Sun Yat-sen University Cancer Center, China. Deidentified tissues were collected in accordance with each institution’s review board and informed consent was waived in both institutions. Clinicopathological characteristics, treatments, and outcomes associated with these patients samples were obtained from the tissue bank. IHC staining was performed as previously described [20].

2.3. Western Blot (WB) and Immunoprecipitation (IP)

Whole cell extracts (WCE), subcellular fractionation and WB were performed as described previously [21]. Primary antibodies used were SPT16 (Abcam, 204343), SSRP1 (Biolegend, 609702), α-HSC70 (B6-Sc7298, Santa Cruz Biotechnology), H2A (Abcam, 26350), GAPDH (CST, 8884s), APE1 (Novus Biologicals, NB100–116), α-tubulin (Abcam, 52666) and AcAPE1 [22]. IP with WCE was performed either with APE1 or AcAPE1 antibody or control IgG (Santa Cruz, sc-2003) as previously described [20]. The immunoprecipitated proteins were resolved in SDS-PAGE and immunoblotted with the indicated antibodies.

2.4. Immunofluorescence (IF)

Cells grown on coverslips were fixed with 4% formaldehyde (Sigma-Aldrich) and stained as described previously [21]. Primary antibodies used were anti-APE1 (1:100; Novus Biologicals, NB100–116), anti-AcAPE1 (1:50) [23, 24], SSRP1 (1:50, Abcam 21584), SPT16 (1:50; Abcam, 204343). Images were acquired by a fluorescence microscope with a 63× oil immersion lens (LSM510; Zeiss), and structured illumination microscopy (SIM) was done with Elyra PS.1 microscope (Carl Zeiss).

2.5. Chromatin Immunoprecipitation (ChIP) Assay

ChIP assay was performed as previously described using anti-AcAPE1 and control IgG (Santa Cruz) [20]. From the immunoprecipitated purified DNA, the p21 and DTL promoter regions were amplified (primers: p21 forward 5’-CAGGCTGTGGCTCTGATTGG-3’, reverse 5’-TTCAGAGTAACAGGCTAAGG-3’; DTL forward 5’-TCCTGCAAATTTCCCGCAAC-3’, reverse 5’-GGCTATGGCGAACAGGAACT-3)’ by genomic qPCR in QuantStudio 3 using SYBR green method.

2.6. MTT Assay

Cell viability was evaluated by MTT assay as described previously [25]. The IC₅₀ values were calculated using GraphPad Prism 7 (San Diego, CA). For synergistic effect combination index (CI) was calculated using CompuSyn software (CI < 0.9 indicates synergism, 0.9–1.1 additivity and >1.1 antagonismas) as described before [26].

2.7. Comet Assay

Comet assay was performed following the manufacturer’s protocol (Trevigen, # 4250–050-K). Tail moment was quantitated in at least 50 independent cells for each condition using the OpenComet v1.3.1 software.

2.8. Xenograft studies

All animal experiments were performed with the approval of the UNMC Institutional Animal Care and Use Committee (IACUC). HD-MB03 and ONS-76 cells (1×10⁶ in 100 μl medium with matrigel) were injected subcutaneously over the left and right flanks in athymic nude mice (Charles Rivers, Wilmington, MA). Subcutaneous tumors were allowed to grow for 2–3 weeks before treatments. The mice were divided into four treatment groups (n=5 mice per group) and received treatments every other day for four weeks. The following drugs: cisplatin (2 mg/kg), CBL0137 (30 mg/kg; Selleck Chemicals, LLC.) were injected intraperitoneally. The combination group received both cisplatin and CBL0137. DMSO was given to the control group. Body weight and tumor volume were measured before each treatment. The mice were euthanized at the end of treatment cycles. Antibodies used for Xenograft tumor IHC included AcAPE1, Ki67 (1:500, CST, 9027), SSRP1, Caspase 3 (1:100, Santa Cruz, sc-7272) and TUNEL assay (Abcam, ab206386). Additive or synergistic effect was examined using the online tool SynergyFinder (https://svnergyfinder.fimm.fi).

2.9. Intracranial orthotopic model

For the orthotopic xenograft study, immunocompromised NOD-scid IL2Rgammanull (or NSG) mice, 6–8 weeks old, were purchased from Jackson Laboratory. HD-MB03 cells expressing RFP/luciferase (2.5 × 10⁵) were suspended in 3 μl neural stem cell media and injected into cerebella (2 mm posterior to the lambda suture, 2 mm lateral to the midline, and 2.5 mm deep) of mice brain using Hamilton neurosyringe (33G) attached to the stereotaxic apparatus (David Kopf Instruments). Mice bearing HD-MB03/Luciferase xenografts were detected by bioluminescence imaging, performed at 5 days after intracranial implantation. Mice with a detectable signal were divided into four groups (n=5): vehicle (DMSO), cisplatin (2 mg/kg), CBL0137 (50 mg/kg), and combination of both cisplatin plus CBL0137, and treatments were given intraperitoneally every four days for 16 days. Mice were imaged on the IVIS2000 system (Perkin-Elmer) at day 5 and at day 21 after injecting with D-luciferin (70 mg/kg). Mice from each group were sacrificed on day 21 and their brains were harvested. H&E staining and IHC analyses of Ki67 were performed to confirm the histology and proliferating tumor cells.

2.10. RNA-seq and qPCR analysis:

HD-MB03 cells were treated with CBL0137 (2 μM) for 30 minutes and then total RNA was isolated. An RNA library was made in our next generation sequencing (NGS) lab at UNMC for RNA-seq analysis. The Tuxedo pipeline [27] was used to find differentially expressed genes (DEGs). The TopHat software was used to align the reads to the human genome. Cufflinks was used to determine FPKM (Fragments per Kilobases of Transcripts per Million mapped reads) values for each transcript in each sample. For validation of OTX2 and MYC gene expression, qPCR were performed using primer pairs: OTX2 forward 5’-GACCACTTCGGTATGGACT-3’ and reverse 5’-TGGACAAGGGATCTGACAGT-3’, and MYC forward5’-CAGGCAGACACATCTCAGGG-3’ and reverse CGTATACTTGGAGAGCGCGT.

2.11. Statistical analysis

Results are shown as the mean ± SEM of three independent experiments. For comparison among multiple groups, one-way analysis of variance (ANOVA) followed by Dunnett’s post-hoc test or Tukey’s HSD test was used depending on the nature of comparison. If data did not distribute normally, Kruskal-Wallis test followed by Games-Howell post-hoc test was used as a non-parametric counterpart of ANOVA for multiple comparison. Statistical analyses were performed in SPSS v22.0 (IBM SPSS Statistics). A p-value of less than 0.05 is considered statistically significant. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

3. Results

3.1. AcAPE1 and SSRP1 levels are elevated in MB patients

We previously showed that APE1 is acetylated (AcAPE1) at AP site damage in chromatin by p300, and modulates its DNA damage repair and transcription regulatory functions [21, 22]. Further, we demonstrated that increased levels of AcAPE1 in colon, lung and pancreatic cancer modulates DNA damage repair [23]. Here, we examined AcAPE1 and SSRP1 levels in a cohort of 71 MB patients’ tissue specimens (44 children and 27 adults) by IHC staining. Clinicopathological characteristics and survival of these patients are shown in Table 1. We found that AcAPE1 and SSRP1 were significantly elevated in MB tissues (Fig. 1A). MB tissues from all histology groups had a higher percentage of AcAPE1 and SSRP1 positive staining (Fig. 1B & 1C). Furthermore, a positive correlation between AcAPE1 and SSRP1 levels (r=0.57, p<0.0001, Fig. 1D) was observed. Kaplan-Meier survival analysis shows that a high level of both AcAPE1 and SSRP1 is associated with poor prognosis (Fig. 1E & 1F). Together, these data suggest that AcAPE1 and FACT are elevated in MB, and levels of AcAPE1 and SSRP1 could serve as a prognostic marker for MB.

Table 1.

Clinical characteristics of the medulloblastoma cohort.

Characteristics	Classical (n=43)	D/N (n=19)	MBEN (n=5)	LCA (n=4)
Age (median, range)	8.5 (2–35)	13 (4–35)	2 (0.5–38)	22.5 (3–33)
Age
≤18	29	10	4	1
>18	14	9	1	3
Sex
Male	25	11	3	4
Female	18	8	2	0
Tumor location
Midline	18	6	1	2
Hemisphere	10	5	1	2
V4	15	8	3	0
Tumor size
≥4 cm	25	14	3	4
<4 cm	14	5	2	0
V4 floor involvement	15	5	2	2
Yes	28	12	3	2
No
Surgical resection
GTR	25	14	4	3
STR	18	5	1	1
Recurrence
Yes	8	3	1	2
No	35	16	4	2
PFS (median, range, months)	12 (1–168)	29 (1–120)	29 (3–30)	8 (1–14)
OS (median, range, months)	15 (1–168)	33 (1–120)	29 (3–30)	15 (1–21)
Death
Yes	7	4	1	2
No	36	15	4	2

Abbreviations: D/N, desmoplastic-nodular; MBEN, medulloblastoma with extensive nodularity; LCA, large cell/anaplastic; V4, fourth ventricle; GTR, gross-total resection; STR, subtotal resection; PFS, progression free survival; OS, overall survival.

Fig 1. AcAPE1 and SSRP1 levels are elevated in MB and is associated with overall patients’ survival.

A) Immunohistochemical staining with AcAPE1 and SSRP1 antibodies was performed in a cohort of 71 MB samples. B & C) The percentage of cells positive for AcAPE1 or SSRP1 from ten random high field in each sample was pooled. Median with 75% and 25% quartile were shown. D) The association between AcAPE1 and SSRP1 expression was analyzed using linear regression. E & F) The overall survival of MB patients in relation to AcAPE1 and SSRP1 expression was analyzed by Kaplan-Meier analysis.

3.2. AcAPE1 colocalizes with FACT in response to radiation and cisplatin treatment

As both FACT and AcAPE1 levels are elevated in MB, we examined whether APE1 interacts with the FACT complex in MB cell lines. We first examined the expression levels of the FACT complex (SSRP1 and SPT16), AcAPE1, and APE1 in different MB cell lines. We found that the expression levels of the FACT complex and AcAPE1 were significantly elevated in MB cell lines as compared to immortalized fetal glial SVG p12 cells (Fig. 2A & 2B). Next, we immunoprecipitated endogenous APE1 from ONS-76, HD-MB03 and DAOY cells and observed that both subunits of FACT are present in the APE1 immunocomplex (Fig. 2C & S1). Confocal microscopy and Pearson correlation coefficient analysis revealed a moderate colocalization of AcAPE1 with SPT16 and SSRP1 in the nucleus of HD-MB03 cells (Fig. 2D & S2). However, we observed that radiation or cisplatin, which both induce DNA damage in the genome, enhanced AcAPE1, SSRP1 and SPT16 levels and their colocalization in cells (Fig. 2E & S3 and Fig. 2F & S4). We also observed increased SPT16 and SSRP1 proteins levels in AcAPE1 immunocomplex from radiation-treated WCEs (Fig. 2G). Together, these data indicate that radiation and cisplatin induce interaction of AcAPE1 with the FACT complex in MB cells.

Fig 2. AcAPE1 interacts and co-localizes with FACT complex in response to cisplatin and radiation.

A & B) The expression levels of SPT16, SSRP1, AcAPE1 and APE1 were determined using Western blot in multiple MB cells lines. C) Coimmunoprecipitation using APE1 antibody was performed with ONS-76 whole cell extracts and immune-complexes were probed with SPT16 and SSRP1 antibodies. D) The expression and co-localization of SSRP1, SPT16 and AcAPE1 were examined using confocal microscopy in HD-MB03 cells. Bar = 5 μm. E & F) Co-localization of SSRP1, SPT16 and AcAPE1 were examined before and after cisplatin (1 μM) and irradiation (5 Gy) treatment. Bar = 10 μm. G). HD-MB03 cells were either left untreated or irradiated and whole cell extracts were immunoprecipitated using AcAPE1 antibody and immunoblotted with SSRP1 and SPT16 antibodies.

3.3. FACT facilitates the binding and acetylation of APE1 at damage sites in chromatin

Previously, we showed that APE1 is acetylated at damage sites in chromatin in lung cancer cells [21]. To examine whether FACT facilitates the recruitment and subsequent acetylation of APE1 at damage sites in chromatin in MB cells, we used siRNA to downregulate SPT16 and SSRP1 and analyzed in WB (Fig. 3A). Confocal microscopy revealed that while cisplatin treatment resulted in an increase in staining of AcAPE1, such effect was abrogated by downregulation of FACT complex (Fig. 3B & S5). Neutral comet assay, which detects DNA DSBs, was used to test whether the absence of the FACT complex affects the repair of radiation-induced damage in cells. In control and FACT knockdown groups, irradiation induced a significant amount of DNA damage as evident by the increased tail moment. However, at 24-hour release, control cells were able to repair the damaged DNA while the FACT downregulated cells were unable to repair DSBs in DNA (Fig. 3C & 3D). ChIP assays in Fig. 3E & 3F showed significant reductions in the occupancy/binding of AcAPE1 in p21 and DTL promoter regions upon radiation damage in FACT downregulated cells. Together, these data provide evidence that the FACT complex promotes the recruitment/binding and subsequent acetylation of APE1 and facilitates damage repair in chromatin.

Fig 3. FACT facilitates acetylation and binding of APE1 to damaged chromatin.

A) FACT complex was depleted by SPT16 and SSRP1 siRNA and expression of SPT16, SSRP1, AcAPE1 and APE1 was examined by Western blot. Bar graph shows the fold change of protein expression. B) The expression of SSRP1 and AcAPE1 level were examined by confocal microscopy after treatment with either cisplatin or FACT siRNA or in combination, Bar = 5 μm. C) The cells were transfected with FACT siRNAs for 48 hours and were then exposed to 5 Gy irradiation. Neutral comet assay was performed immediately and at 24-hour after irradiation. D) Tail moment of 50 cells was examined in each treatment group and the bar graph was shown. E & F) The binding/occupancy of AcAPE1 to p21 and DTL promoter regions were examined using ChIP-qPCR assay.

3.4. Inhibition of FACT with CBL0137 delayed radiation-induced DNA damage repair

To further examine if FACT functionally cooperates with APE1 and facilitates the recruitment and acetylation of APE1 at damage sites in chromatin, we used the small molecule CBL0137, a second-generation CX and potent inhibitor of the FACT complex [15]. It was shown that CBL0137 induces redistribution of FACT from nucleus to chromatin, and FACT binds to unfolded nucleosome and trapped in chromatin in the presence of CBL0137 [15, 28]. Treatment of cells with CBL0137 decreased the soluble SSRP1 and SPT16 proteins in WCE in a dose-dependent manner in HD-MB03 and ONS-76 cells (Fig. 4A). As previously reported, we also observed a reduction in the SSRP1 and SPT16 levels in nuclear extracts with a concurrent increased of their levels in chromatin fractions after CBL0137 treatment [29, 30] (Fig. 4B & 4C). We found that inhibition of FACT with CBL0137 decreased the acetylation of APE1 levels without altering the total APE1 levels in cells (Fig. 4D, S6–S7). To provide evidence for the role of FACT in facilitating repair of radiation-induced DNA damage, we used a single-cell alkaline comet assay, which detects both SSBs and DSBs. Control cells exhibited tails indicative of DNA damage at 6 hours after radiation exposure but were able to recover by 36 hours. However, CBL0137 treated cells retained significantly more comet tails even after 36 hours of recovery (Fig. 4E & 4F). Together, these findings suggest that CBL0137 inhibits APE1 acetylation, and in the presence of CBL0137, cells cannot efficiently repair radiation-induced DNA damage.

Fig 4. CBL0137 (CX) inhibits FACT complex and reduces acetylation of APE1, and treatment of CX sensitizes MB cells to chemo and radiation.

A-C) Cells were treated with indicated doses of CBL0137 (CX). Whole cell extracts (A), soluble nuclear fraction and chromatin extracts (B &C) were used to examine the expression of SPT16 and SSRP1. D) HD-MB03 cells were treated with CX and confocal microscopy was performed with APE1 and AcAPE1. Bar = 10 μm. E) Cells were either left untreated or irradiated (5 Gy) with and without CX. Alkaline comet assay was used to detect DNA damage at 6 hour and 36-hour after irradiation. F) Tail moment of 50 cells was examined in each treatment group and the bar graph was shown. G-H) HD-MB03 and ONS-76 cells were pretreated with CX and exposed to indicated doses of cisplatin (G) or irradiation (H) and MTT assay was performed. I-L) Cells were pretreated with CX and exposed to cisplatin (I) or irradiation (J). The effect was examined using colony formation assays. Representative images of colonies were shown (K & L).

3.5. Curaxin/CBL0137 treatment sensitizes MB cell lines to cisplatin and irradiation

We examined whether targeting FACT with CBL0137 enhanced the efficacy of radiation and cisplatin in MB cells. To determine whether CBL0137 alone induces DNA damage, cells were first treated with a range of CBL0137 doses. We found that treatment with 2 μM CBL0137 alone had a minimal effect (20% reduction) on cell viability and colony formation (Fig. S8A & S8B). However, pre-treatment of cells with 2 μM CBL0137 for 1 hour followed by radiation and cisplatin treatment reduced cell viability in a dose-dependent manner (Fig. 4G & 4H). We found ~10 -fold and ~6 -fold decrease in the IC₅₀ of cisplatin in HD-MB03 and in ONS-76 cells, respectively, when pretreated with CBL0137 with combination index (CI) of 0.9 and 0.2. Similarly, we also observed ~2.7 -fold and ~2 -fold decrease in IC₅₀ of radiation with CI of 0.23 and 0.05 in HD-MB03 and ONS-76 cells, respectively. The CI suggests that combination of CBL0137 with cisplatin or radiation demonstrates synergistic effects (< 0.9). In addition, colony formation assays revealed that pretreating cells with 2 μM CBL0137 increased their sensitivity to cisplatin and irradiation (Fig. 4I–4L). Together, these data suggest that a combination of CBL0137 and cisplatin or radiation has significant synergistic effects [26] on MB cell survival.

3.6. Combination of CBL0137 and cisplatin suppresses MB tumor growth in both subcutaneous and intracranial orthotopic xenograft models

To examine whether inhibition of FACT with CBL0137 sensitizes MB cells to cisplatin and inhibits tumor growth in vivo, first we utilized subcutaneous tumor xenograft model with HD-MB03 (Fig. 5A & 5B) and ONS-76 (Fig. S9A & S9B) cells. Tumor growth curve analysis showed that single-agent treatment either with CBL0137 or cisplatin alone had a moderate effect on tumor growth compared to the vehicle group; however, the combination of CBL0137 and cisplatin significantly inhibited tumor growth. The combination of cisplatin with CBL0137 was well tolerated at the scheduled doses as there was no significant difference in weight loss in mice among groups (Fig. S9C). Moreover, histologic examination of major organs, including liver, lung, spleen, heart, and kidneys, did not demonstrate toxicity after completion of the treatment (Fig. S9D). Further, IHC analysis of Ki-67, TUNEL, and caspase-3 showed that combination treatment suppressed cell proliferation (Fig. 5C) and increased the number of apoptotic cells. Consistent with in vitro results, CBL0137 treatment decreased the staining intensity of AcAPE1 in these tumor sections. Together, these data suggest that CBL0137 augments cisplatin treatment in vivo, and the combination decreases MB tumor burden and cellular proliferation, while increasing apoptosis.

Fig 5. Combination of CBL0137 and cisplatin suppresses xenograft MB tumor growth in vivo.

A) HD-MB03 cells were subcutaneously implanted in nude mice and resected xenograft tumor after completion of treatment were shown. B) Tumor volume was measured at indicated days and tumor growth curve was plotted. C) The xenograft sections were submitted for TUNEL assay and IHC analysis of Ki-67, caspase-3, SSRP1 and AcAPE1.

We next tested whether combination of CBL0137 and cisplatin treatment suppresses MB growth in an intracranial orthotopic xenograft model. HDMB-03 cells stably expressing RFP/luciferase were injected into the cerebella of NSG mice and treated with vehicle (DMSO), CBL0137, cisplatin alone or combination of CBL0137 plus cisplatin. Five days after implantation, drugs were injected every four days for 16 days, and bioluminescence IVIS imaging were performed. IVIS imaging at day 5 and at day 21 (post-implantation) of these mice from each group show that while treatment with either CBL0137 or cisplatin alone had a moderate effect in suppression of tumor growth, combination of CBL0137 plus cisplatin significantly suppressed tumor growth in vivo (Fig. 6A). These mice were sacrificed after the final treatment and their brains were harvested. As shown in Fig. 6B, H&E staining and IHC analysis of Ki67 revealed a much larger tumor at the cerebellum region of the vehicle-treated mouse compared to the CBL0137 or cisplatin treated mice. Further, a significant suppression of tumor growth was observed in CBL0137 plus cisplatin treated mouse, suggesting that CBL0137 augments cisplatin treatment, and the combination effectively suppresses MB tumor growth in an orthotopic xenograft model.

Fig. 6. CBL0137 plus cisplatin effectively inhibits the growth of MB in intracranial orthotopic tumor model.

A) RFP/Luciferase expressing HD-MB03 cells were implanted into cerebella of NSG mice for tumor growth and the mice were treated with the indicated drugs. IVIS images of representative mouse from each treatment group at day 5 (treatment started), and at day 21 (mice sacrificed) after implantation are shown. B) Gross appearance of the whole brain of the vehicle and the drugs treated mice are shown. H&E staining and IHC staining of Ki67 of section of mice brain show the tumor and proliferating cells.

3.7. CBL0137 alters expression of genes that promote cell proliferation and chemoresistance

Previous studies showed that CBL0137 alters gene expression by targeting FACT [17, 18]. We examined the possibility that CBL0137 modulates tumor growth or sensitivity via altering the expression of genes involved in proliferation or chemoresistance. We performed RNA-seq analysis in Myc-amplified HD-MB03 cells and compared the gene expression profiles before and after treatment of CBL0137. The correlation coefficients R² for duplicates in control and treatment groups were greater than 0.99 (Fig. S10). Differential gene expression analysis revealed that CBL0137 treatment upregulated 1,655 genes and downregulated 1,801 genes (Fig. 7A). Consistently, gene ontology analysis showed that genes significantly affected by CBL0137 were involved in the regulation of the cell cycle, phase transition, spliceosome, RNA transport, ribosomes, and p53 signaling pathway (Fig. 7B). We found alterations of the expression of multiple genes that are known to promote MB proliferation and chemoresistance in CBL0137 treated cells, including SLC7A11 [31], MT2A [32], NRL [33], c-MYC [34], and OTX2 [35] (only the top 20 affected genes were shown, Fig. 7C). We further validated, by real-time PCR analysis, that CBL0137 treatment downregulated proto-oncogenes MYC and OTX2 (Fig. 7D) which are known to promote oncogenesis and chemoresistance in group 3 MB [36, 37]. Together, these data suggest that the FACT inhibitor CBL0137 exerts a broad effect on the expression of multiple genes which are involved in progression of MB.

Fig 7. CBL0137 (CX) alters expression of genes that promote cell proliferation and chemoresistance.

A) Differential gene expression analysis was used to identify genes that were significantly upregulated or downregulated by CX treatment. B) Scatter plot of GO enrichment of differentially expressed genes was shown. C) A heatmap of the top 20 differentially expressed genes was shown. D) The effect of CX on the mRNA expression of c-MYC and OTX2 was examined by qPCR.

4. Discussion

Radiation and platinum-based chemotherapy are effective against a variety of pediatric cancers including MB. Unfortunately, administration of cisplatin is complicated by prominent dose-limiting neurotoxicity and progressive hearing loss which affect the quality of life of childhood cancer survivors. Therefore, development of novel therapeutic approaches by delineating specific molecular targets are highly desirable to increase the efficacy of chemo-radiation and improve therapy-related toxicity and outcomes in high-risk pediatric patients.

In this study, we have demonstrated that targeting FACT complex with CBL0137 significantly improves the efficacy of chemo and radiation in vitro and in vivo in a preclinical mouse model. We have shown that FACT complex functionally interacts with APE1 and facilitates the recruitment and acetylation of APE1 at damage sites to facilitate DNA repair. We further demonstrated that FACT inhibitor CBL0137 inhibits DNA damage repair and alters expression of oncogenes including MYC and OTX2, highlighting the mechanisms by which CBL0137 sensitizes MB cells to chemoradiation. Our study suggests that targeting FACT with CBL0137 is a promising strategy to sensitize MB cells to chemoradiation, and CBL0137 could be used as an adjuvant therapy to lower the side effects in MB patient.

Several earlier studies demonstrated that APE1 is overexpressed in multiple types of cancer including MB, non-small cell lung cancer, colon, pancreatic cancer, and overexpression of APE1 is associated with patients’ resistance to chemotherapy and poor prognosis [6, 38]. APE1 primarily repairs AP sites or SSB that are generated after treatment with radiation and many chemotherapeutic drugs including cisplatin [5]. The repair mechanism by which APE1 acts on naked DNA or nucleosomal DNA substrate has been extensively investigated in vitro [5]. However, eukaryotic cells must repair DNA lesions within the context of nucleosome in chromatin. We previously showed that APE1 is acetylated at multiple Lys residues and acetylation which occurs after APE1’s chromatin binding enhances its AP-endonuclease (repair activity) [21]. Here, we have shown that nucleosome remodeling histone chaperone FACT complex facilitates the recruitment of APE1 to damage sites and its subsequent acetylation in chromatin. FACT complex originally discovered as histone chaperone that interacts with RNA polymerases [39, 40] and facilitates transcription by disrupting nucleosomes in their path and by aiding in the removal and re-deposition of histones during transcriptional elongation [12, 41]. However, increasing evidence suggests that FACT is involved in multiple DNA damage repair pathways. It has been shown that FACT is present in active transcription-coupled NER repair complexes, and its SPT16 subunit is involved in eviction and deposition of H2A and H2B at sites of UV-induced DNA damage [13]. FACT has been shown to act in concert with RSC to facilitate excision of DNA lesions during the initial step of BER [42]. However, its role in promoting drug resistance in cancer has not been investigated. Here, we found that downregulation of FACT or inhibition of its function by CBL0137 affects DNA damage repair and expression of a subset of genes, and sensitizes tumor cells to radiation and cisplatin. Thus, our study demonstrates the interplay between a histone chaperone complex and DNA repair pathways and transcription in promoting drug resistance in tumor cells.

Like APE1, elevated FACT expression was also found in clinical tumor samples compared to corresponding normal tissues [17, 28]. IHC staining of SSRP1 on tissue microarrays containing primary and metastatic tumors of different types including invasive breast ductal and lobular carcinoma, non-small cell lung cancer (NSCLC), renal cell carcinoma, prostatic adenocarcinoma, pancreatic ductal adenocarcinoma, and glioblastoma revealed elevated levels of SSRP staining in tumors [17, 28]. The most striking correlation between high FACT levels and poor survival was observed in pediatric neuroblastoma [29]. Recently, our studies also demonstrated elevated levels of FACT complex in colon cancer and bladder cancer patients’ tissue samples and their association with patients prognosis [43, 44]. In this study, using 71 patients’ MB cohort we demonstrate that both AcAPE1 and FACT complex are also elevated in MB, and their overexpression are associated with poor survival. While APE1 can be targeted, FACT complex is a more appealing target for the following reasons. First, FACT complex is involved in multiple DNA repair pathways as well as regulating transcription. Targeting FACT simultaneously inhibits multiple DNA damage repair pathways and alters gene expression [13, 14, 17, 18]. Second, FACT complex is undetectable in normal cells in mammalian tissues, except for undifferentiated and stem-like cells [28]. It is upregulated during in vitro transformation and promotes survival and growth of established tumor cells [28]. Such differential expression among normal and tumor tissues is advantageous in lowering side effects when it comes to treatment. Third, there is no available molecule that targets APE1 DNA repair function with high efficacy in vivo [45], but there is a readily available anticancer small molecule, CBL0137, which effectively inhibits FACT functions in cells and is under clinical trial for other solid tumors [46, 47].

Curaxin is a group of small molecules that were originally identified to simultaneously activate p53 and inhibit NF-κB without causing detectable genotoxicity [48]. The second generation curaxin, CBL0137, binds DNA without causing detectable DNA damage and leads to nucleosome unfolding [16]. This results in opening up FACT-binding sites in histone H2A and H2B, which are normally hidden inside the nucleosome where they trap FACT in chromatin [16, 49]. Consistent with earlier observation, our data shows CBL0137 treatment reduced soluble FACT and increased trapping of FACT complex, which likely affects its normal histone removal and/or deposition functions during DNA repair and transcription elongation [29]. Supporting this, our data show that downregulation of FACT complex or inhibiting it with CBL0137 both significantly reduced radiation-induced damage repair and altered gene expression. Importantly, using both subcutaneous and intracranial orthotopic xenograft models we have demonstrated that while cisplatin or CBL0137 treatment alone had a moderate effect on suppression of MB tumor growth combination of cisplatin and CBL0137 significantly suppresses MB growth in vivo. Although a recent study by Wang et al. [19] also showed that CBL0137 suppresses group 3 MB growth in vivo, our study demonstrates that CBL0137 not only suppress MB growth but also augmented the efficacy of standard chemotherapy cisplatin in vivo, highlighting the potential of CBL0137 as an adjuvant therapy to sensitize MB and to lower the side effects of standard chemotherapy. FACT inhibitor CBL0137 has been shown to synergize with cisplatin in small cell lung cancer by increasing NOTCH1 expression and targeting cancer stem cells [18]. Our RNA seq data also revealed upregulation of NOTCH 3 expression and downregulated stem cell marker SOX2. FACT expression was also shown to be elevated in glioblastoma (GBM) and is positively correlated with GBM stem cells marker expression [17]. Nonetheless, here we add to the current knowledge by providing compelling evidence that CBL0137 exhibits strong synergy with cisplatin and radiation in killing MB cells by inhibiting DNA damage repair and altering gene expression. Our data show that treatment of CBL0137 alters gene expression in MYC-amplified group 3 MB cells, the most aggressive group of pediatric MB that is often resistant to radiation and chemotherapy. CBL0137 significantly downregulated MYC and OTX gene expression in HD-MB03 cells. Like MYC, OTX2 is highly expressed in all group 3 MB and is amplified in 8% of cases, making it the second major amplified gene after MYC. A previous study showed that OTX2 is a major activator of genes involved in cell growth and survival. Consistent with our results, a recently published transcriptome analyses in MB cells also showed that CBL0137 preferentially suppressed cell-cycle and DNA-repair related biological processes [19]. CBL0137 has received increasing attention because it has moved to phase II clinical trials for patients with hematological malignancies, metastatic melanoma, or sarcoma (NCT02931110 and NCT03727789).

In conclusion, we report for the first time that elevated levels of AcAPE1 and FACT complex in MB are associated with poor prognosis. FACT complex interacts with and facilitates the access of APE1 to damage sites and promotes DNA damage repair. Targeting FACT complex with small molecule CBL0137 significantly improves the efficacy of cisplatin and radiation in vitro and in vivo in a preclinical model. The second generation curaxin CBL0137 represents a highly translatable and targeted therapeutic agent that can be used as an adjuvant therapy to sensitize MB and to lower the side effects of current standard therapy in MB patients while achieving superior treatment efficacy.

Highlights.

1. FACT complex facilitates the binding and acetylation of APE1 at damage sites in chromatin and promotes drug resistance.

2. Elevated levels of AcAPE1 and FACT complex are associated with poor prognosis in MB.

3. FACT inhibitor CBL0137 improves the potency of cisplatin and radiation in MB.