Published papers

Photons, such as X- or γ-rays, induce DNA damage (distributed throughout the nucleus) as a result of low-density energy deposition. In contrast, particle irradiation with high linear energy transfer (LET) deposits high-density energy along the particle track. High-LET heavy-ion irradiation generates a greater number and more complex critical chromosomal aberrations, such as dicentrics and translocations, compared with X-ray or γ irradiation. In addition, the formation of >1000 bp deletions, which is rarely observed after X-ray irradiation, has been identified following high-LET heavy-ion irradiation. Previously, these chromosomal aberrations have been thought to be the result of misrepair of complex DNA lesions, defined as DNA damage through DNA double-strand breaks (DSBs) and single-strand breaks as well as base damage within 1-2 helical turns (<3-4 nm). However, because the scale of complex DNA lesions is less than a few nanometers, the large-scale chromosomal aberrations at a micrometer level cannot be simply explained by complex DNA lesions. Recently, we have demonstrated the existence of clustered DSBs along the particle track through the use of super-resolution microscopy. Furthermore, we have visualized high-level and frequent formation of DSBs at the chromosomal boundary following high-LET heavy-ion irradiation. In this review, we summarize the latest findings regarding the hallmarks of DNA damage structure and the repair pathway following heavy-ion irradiation. Furthermore, we discuss the mechanism through which high-LET heavy-ion irradiation may induce dicentrics, translocations and large deletions.

Programmed death-ligand 1 (PD-L1) is a key factor influencing cancer immunotherapy; however, the regulation of PD-L1 expression in cancer cells remains unclear, particularly regarding DNA damage, repair and its signalling. Herein, we demonstrate that oxidative DNA damage induced by exogenously applied hydrogen peroxide (H2O2) upregulates PD-L1 expression in cancer cells. Further, depletion of the base excision repair (BER) enzyme DNA glycosylase augments PD-L1 upregulation in response to H2O2. PD-L1 upregulation in BER-depleted cells requires ATR/Chk1 kinase activities, demonstrating that PD-L1 upregulation is mediated by DNA damage signalling. Further analysis of The Cancer Genome Atlas revealed that the expression of PD-L1 is negatively correlated with that of the BER/single-strand break repair (SSBR) and tumours with low BER/SSBR gene expression show high microsatellite instability and neoantigen production. Hence, these results suggest that PD-L1 expression is regulated in cancer cells via the DNA damage signalling and neoantigen-interferon-γ pathway under oxidative stress.

Photon radiation therapy is a major curative treatment for cancer. However, the lack of robust predictive biomarkers for radiosensitivity precludes personalized radiation therapy. Clonogenic assays are the gold standard method for measuring the radiosensitivity of cancer cells. Although a large number of publications describe the use of clonogenic assays to measure cancer cell radiosensitivity, the robustness of results from different studies is unclear. To address this, we conducted a comprehensive detailed literature search of 256 common cancer cell lines and identified the eight cell lines most-frequently examined for photon sensitivity using clonogenic assays. Survival endpoints and experimental parameters from all 620 relevant experiments were compiled and analyzed. We found that the coefficients of variation for SF2 (surviving fraction after 2 Gy irradiation) and for D10 (dose that yields a surviving fraction of 10%) were below 30% for all cell lines, indicating that SF2 and D10 have acceptable inter-assay precision. These data support further analysis of published data on clonogenic assays using SF2 and D10 as survival endpoints, which facilitates robust identification of biological profiles representative of cancer cell sensitivity to photons.

Radiotherapy is an essential component of cancer therapy. Carbon ion radiotherapy (CIRT) promises to improve outcomes compared with standard of care in many cancers. Nevertheless, clinicians often observe in-field recurrence after CIRT. This indicates the presence of a subset of cancers that harbor intrinsic resistance to CIRT. Thus, the development of methods to identify and sensitize CIRT-resistant cancers is needed. To address this issue, we analyzed a unique donor-matched pair of clinical specimens: a treatment-naïve tumor, and the tumor that recurred locally after CIRT in the same patient. Exon sequencing of 409 cancer-related genes identified enrichment of somatic mutations in FGFR3 and FGFR4 in the recurrent tumor compared with the treatment-naïve tumor, indicating a pivotal role for FGFR signaling in cancer cell survival through CIRT. Inhibition of FGFR using the clinically available pan-FGFR inhibitor LY2874455 sensitized multiple cancer cell lines to carbon ions at 3 Gy (RBE: relative biological effectiveness), the daily dose prescribed to the patient. The sensitizer enhancement ratio was 1.66 ± 0.17, 1.27 ± 0.09, and 1.20 ± 0.18 in A549, H1299, and H1703 cells, respectively. Our data indicate the potential usefulness of the analytical pipeline employed in this pilot study to identify targetable mutations associated with resistance to CIRT, and of LY21874455 as a sensitizer for CIRT-resistant cancers. The results warrant validation in larger cohorts.

BACKGROUND AND PURPOSE: Integrated analysis of existing radiosensitivity data obtained by the gold-standard clonogenic assay has the potential to improve our understanding of cancer cell radioresistance. However, extraction of radiosensitivity data from the literature is highly labor-intensive. To aid in this task, using deep convolutional neural networks (CNNs) and other computer technologies, we developed an analysis pipeline that extracts radiosensitivity data derived from clonogenic assays from the literature. MATERIALS AND METHODS: Three classifiers (C1-3) were developed to identify publications containing radiosensitivity data derived from clonogenic assays. C1 uses Faster Regions CNN with Inception Resnet v2 (fRCNN-IRv2), VGG-16, and Optical Character Recognition (OCR) to identify publications that contain semi-logarithmic graphs showing radiosensitivity data derived from clonogenic assays. C2 uses fRCNN-IRv2 and OCR to identify publications that contain bar graphs showing radiosensitivity data derived from clonogenic assays. C3 is a program that identifies publications containing keywords related to radiosensitivity data derived from clonogenic assays. A program (iSF2) was developed using Mask RCNN and OCR to extract surviving fraction after 2-Gy irradiation (SF2) as assessed by clonogenic assays, presented in semi-logarithmic graphs. The efficacy of C1-3 and iSF2 was tested using seven datasets (1805 and 222 publications in total, respectively). RESULTS: C1-3 yielded sensitivity of 91.2% ± 3.4% and specificity of 90.7% ± 3.6%. iSF2 returned SF2 values that were within 2.9% ± 2.6% of the SF2 values determined by radiation oncologists. CONCLUSION: Our analysis pipeline is potentially useful to acquire radiosensitivity data derived from clonogenic assays from the literature.

Ribosomes are important cellular components that maintain cellular homeostasis through overall protein synthesis. The nucleolus is a prominent subnuclear structure that contains ribosomal DNA (rDNA) encoding ribosomal RNA (rRNA), an essential component of ribosomes. Despite the significant role of the rDNA‑rRNA‑ribosome axis in cellular homeostasis, the stability of rDNA in the context of the DNA damage response has not been fully investigated. In the present study, the number and morphological changes of nucleolin, a marker of the nucleolus, were examined following ionizing radiation (IR) in order to investigate the impact of DNA damage on nucleolar stability. An increase in the number of nucleoli per cell was found in HCT116 and U2OS cells following IR. Interestingly, the IR‑dependent increase in nucleolar fragmentation was enhanced by p53 deficiency. In addition, the morphological analysis revealed several distinct types of nucleolar fragmentation following IR. The pattern of nucleolar morphology differed between HCT116 and U2OS cells, and the p53 deficiency altered the pattern of nucleolar morphology. Finally, a significant decrease in rRNA synthesis was observed in HCT116 p53‑/‑ cells following IR, suggesting that severe nucleolar fragmentation downregulates rRNA transcription. The findings of the present study suggest that p53 plays a key role in protecting the transcriptional activity of rDNA in response to DNA damage.

Chemical modifications and adducts at DNA double-strand break (DSB) ends must be cleaned before re-joining by non-homologous end-joining (NHEJ). MRE11 nuclease is essential for efficient removal of Topoisomerase II (TOP2)-DNA adducts from TOP2 poison-induced DSBs. However, mechanisms in MRE11 recruitment to DSB sites in G1 phase remain poorly understood. Here, we report that TOP2-DNA adducts are expeditiously removed through UBC13-mediated polyubiquitination, which promotes DSB resection in G2 phase. We found that this ubiquitin signaling is required for efficient recruitment of MRE11 onto DSB sites in G1 by facilitating localization of RAP80 and BRCA1 to DSB sites and complex formation between BRCA1 and MRE11 at DSB sites. UBC13 and MRE11 are dispensable for restriction-enzyme-induced “clean” DSBs repair but responsible for over 50% and 70% of NHEJ-dependent repair of γ-ray-induced “dirty” DSBs, respectively. In conclusion, ubiquitin signaling promotes nucleolytic removal of DSB blocking adducts by MRE11 before NHEJ.

Carbon-ion radiotherapy (CIRT) has strong antitumor effects and excellent dose conformity. In head-and-neck squamous cell carcinoma (HNSCC), human papillomavirus (HPV) status is a prognostic factor for photon radiotherapy outcomes. However, the effect of HPV status on the sensitivity of HNSCCs to carbon ions remains unclear. Here, we showed that the relative biological effectiveness (RBE) of carbon ions over X-rays was higher in HPV-negative cells than in HSGc-C5 cells, which are used for CIRT dose establishment, whereas the RBE in HPV-positive cells was modest. These data indicate that CIRT is more advantageous in HPV-negative than in HPV-positive HNSCCs.

Abstract Radiotherapy treatment strategies should be personalized based on the radiosensitivity of individual tumors. Clonogenic assays are the gold standard method for in vitro assessment of radiosensitivity. Reproducibility is the critical factor for scientific rigor; however, this is reduced by insufficient reporting of methodologies. In reality, the reporting standards of methodologies pertaining to clonogenic assays remain unclear. To address this, we performed a literature search and qualitative analysis of the reporting of methodologies pertaining to clonogenic assays. A comprehensive literature review identified 1672 papers that report the radiosensitivity of human cancer cells based on clonogenic assays. From the identified papers, important experimental parameters (i.e. number of biological replicates, technical replicates, radiation source and dose rate) were recorded and analyzed. We found that, among the studies, (i) 30.5% did not report biological or technical replicates; (ii) 47.0% did not use biological or technical replicates; (iii) 3.8% did not report the radiation source; and (iv) 32.3% did not report the dose rate. These data suggest that reporting of methodologies pertaining to clonogenic assays in a considerable number of previously published studies is insufficient, thereby threatening reproducibility. This highlights the need to raise awareness of standardization of the methodologies used to conduct clonogenic assays.

Micronuclei (MN) trigger antitumor immune responses via the cyclic GMP-AMP synthase-signaling effector stimulator of interferon genes (cGAS-STING) pathway. Radiotherapy induces MN in peripheral blood lymphocytes. However, data for solid tumors are lacking. Here, we analyzed MN post-radiotherapy in solid tumor samples. Tumor biopsy specimens were obtained from seven prospectively recruited patients with cervical cancer, before treatment and after receiving radiotherapy at a dose of 10 Gy (in five fractions). The samples were stained with 4′,6-diamidino-2-phenylindole dihydrochloride, and 200 nuclei per sample were randomly identified and assessed for the presence of MN or apoptosis, based on nuclear morphology. The median number of MN-harboring nuclei was significantly greater in samples from patients treated with radiotherapy than in pre-treatment samples (151 (range, 16–327) versus 28 (range, 0–61); p = 0.015). No significant differences in the number of apoptotic nuclei were observed between pre-treatment and 10 Gy samples (5 (range, 0–30) versus 12 (range, 2–30); p = 0.30). This is the first report to demonstrate MN induction by radiotherapy in solid tumors. The results provide clinical evidence of the activation of antitumor immune responses by radiotherapy.

Cancer therapy using immune checkpoint inhibitors (ICIs) is a promising clinical strategy for patients with multiple types of cancer. The expression of programmed cell death ligand-1 (PD-L1), an immune-suppressor ligand, in cancer cells is a factor that influences the efficacy of ICI therapy, particularly in the anti-programmed cell death protein-1 (PD-1)/PD-L1 antibody therapy. PD-L1 expression in cancer cells are associated with tumor mutation burden including microsatellite instability because the accumulation of mutations in the cancer genome can produce abnormal proteins via mutant mRNAs, resulting in neoantigen production and HLA-neoantigen complex presentation in cancer cells. HLA-neoantigen presentation promotes immune activity within tumor environment; therefore, known as hot tumor. Thus, as the fidelity of DNA repair affects the generation of genomic mutations, the status of DNA repair and signaling in cancer cells can be considered prior to ICI therapy. The Cancer Genome Atlas (TCGA) and The Cancer Immunome Atlas (TCIA) database analysis showed that tumor samples harboring mutations in any non-homologous end joining, homologous recombination, or DNA damage signaling genes exhibit high neoantigen levels. Alternatively, an urgent task is to understand how the DNA damage-associated cancer treatments change the status of immune activity in patients because multiple clinical trials on combination therapy are ongoing. Recent studies demonstrated that multiple pathways regulate PD-L1 expression in cancer cells. Here, we summarize the regulation of the immune response to ICI therapy, including PD-L1 expression, and also discuss the potential strategies to improve the efficacy of ICI therapy for poor responders from the viewpoint of DNA damage response before or after DNA damage-associated cancer treatment.

Clonogenic assays are the gold standard to measure in vitro radiosensitivity, which use two cell plating methods, before or after irradiation (IR). However, the effect of the plating method on the experimental outcome remains unelucidated. By using common cancer cell lines, here we demonstrate that pre-IR and post-IR plating methods have a negligible effect on the clonogenic assay-derived photon sensitivity as assessed by SF2, SF4, SF6, SF8, D10, or D50 (N.B. SFx indicates the survival at X Gy; Dx indicates the dose providing X% survival). These data provide important biological insight that supports inter-study comparison and integrated analysis of published clonogenic assay data regardless of the plating method used.

AbstractTo obtain quantitative volumetric data for the Golgi apparatus after ionizing radiation (IR) using super-resolution three-dimensional structured illumination (3D-SIM) microscopy. Normal human retinal pigment epithelial (RPE) cells were irradiated with X-rays (10 Gy), followed by immunofluorescence staining of the Golgi marker RCAS1. 3D-SIM imaging was performed using DeltaVision OMX version 4 and SoftWoRx 6.1. Polygon rendering and spot signal identification were performed using Imaris 8.1.2. Differences between groups were assessed by Welch’s t test. RCAS1 signals in untreated cells were located adjacent to nuclei and showed a reticular morphology. Upon IR, the area of RCAS1 signals expanded while retaining the reticular morphology. Polygon rendering imaging revealed that the volume of RCAS1 at 48 h post-IR was greater than that for unirradiated cells (93.7 ± 19.0 μm3 vs. 33.0 ± 4.2 μm3, respectively; P < 0.001): a 2.8-fold increase. Spot signal imaging showed that the number of RCAS1 spot signals post-IR was greater than that for unirradiated cells [3.4 ± 0.8 (× 103) versus 1.3 ± 0.2 (× 103), respectively; P < 0.001]: a 2.7-fold increase. This is the first study to report quantitative volumetric data of the Golgi apparatus in response to IR using super-resolution 3D-SIM microscopy.

Radiotherapy induces an immune response in the cancer microenvironment that may influence clinical outcome. The present study aimed to analyse the alteration of CD8+ T-cell infiltration and programmed death-ligand 1 (PD-L1) expression following radiotherapy in clinical samples from patients with uterine cervical squamous cell carcinoma. Additionally, the current study sought to analyse the association between these immune responses and clinical outcomes. A total of 75 patients who received either definitive chemoradiotherapy or radiotherapy were retrospectively analyzed. CD8+ T-cell infiltration and PD-L1 expression were determined by immunohistochemistry using biopsy specimens before radiotherapy (pre-RT) and after 10 Gy radiotherapy (post-10 Gy). The PD-L1+ rate was significantly increased from 5% (4/75) pre-RT to 52% (39/75) post-10 Gy (P<0.01). Despite this increase in the PD-L1+ rate post-10 Gy, there was no significant association between both pre-RT and post-10 Gy and overall survival (OS), locoregional control (LC) and progression-free survival (PFS). On the other hand, the CD8+ T-cell infiltration density was significantly decreased for all patients (median, 23.1% pre-RT vs. 16.9% post-10 Gy; P=0.038); however, this tended to increase in patients treated with radiotherapy alone (median, 17.7% pre-RT vs. 24.0% post-10 Gy; P=0.400). Notably, patients with high CD8+ T-cell infiltration either pre-RT or post-10 Gy exhibited positive associations with OS, LC and PFS. Thus, the present analysis suggested that CD8+ T-cell infiltration may be a prognostic biomarker for patients with cervical cancer receiving radiotherapy. Furthermore, immune checkpoint inhibitors may be effective in patients who have received radiotherapy, since radiotherapy upregulated PD-L1 expression in cervical cancer specimens.

Abstract Programmed death ligand 1 (PD-L1) expression on the surface of cancer cells affects the efficacy of anti-PD-1/PD-L1 immune checkpoint therapy. However, the mechanism underlying PD-L1 expression in cancer cells is not fully understood, particularly after ionizing radiation (IR). Here, we examined the impact of high linear energy transfer (LET) carbon-ion irradiation on the expression of PD-L1 in human osteosarcoma U2OS cells. We found that the upregulation of PD-L1 expression after high LET carbon-ion irradiation was greater than that induced by X-rays at the same physical and relative biological effectiveness (RBE) dose, and that the upregulation of PD-L1 induced by high LET carbon-ion irradiation was predominantly dependent on ataxia telangiectasia and Rad3-related (ATR) kinase activity. Moreover, we showed that the downstream signaling, e.g. STAT1 phosphorylation and IRF1 expression, was upregulated to a greater extent after high LET carbon-ion irradiation than X-rays, and that IRF1 upregulation was also ATR dependent. Finally, to visualize PD-L1 molecules on the cell surface in 3D, we applied immunofluorescence-based super-resolution imaging. The three-dimensional structured illumination microscopy (3D-SIM) analyses revealed substantial increases in the number of presented PD-L1 molecules on the cell surface after high LET carbon-ion irradiation compared with X-ray irradiation.

Chromosome rearrangements, which are structural chromosomal abnormalities commonly found in human cancer, result from the misrejoining between two or more DNA double-strand breaks arising at different genomic regions. Consequently, chromosome rearrangements can generate fusion genes that promote tumorigenesis. The mechanisms of chromosome rearrangement have been studied using exogenous double-strand break inducers, such as radiation and nucleases. However, the mechanism underlying the occurrence of chromosome rearrangements in the absence of exogenous double-strand break-inducing stimuli is unclear. This study aimed to identify the major source of chromosome rearrangements and the DNA repair pathway that suppresses them. DNA repair factors that potentially suppress gene fusion were screened using The Cancer Genome Atlas dataset. In total, 22 repair factors whose expression levels were negatively correlated with the frequency of gene fusion were identified. More than 60% of these repair factors are involved in homologous recombination, a major double-strand break repair pathway. We hypothesized that DNA single-strand breaks are the source of double-strand breaks that lead to chromosome rearrangements. This study demonstrated that hydrogen peroxide (H2O2)-induced single-strand breaks gave rise to double-strand breaks in a replication-dependent manner. Additionally, H2O2 induced the formation of RPA and RAD51 foci, which indicated that double-strand breaks derived from single-strand breaks were repaired through homologous recombination. Moreover, treatment with H2O2 promoted the formation of radial chromosomes, a type of chromosome rearrangements, only upon the downregulation of homologous recombination factors, such as BRCA1 and CtIP. Thus, single-strand breaks are the major source of chromosome rearrangements when the expression of homologous recombination factors is downregulated.

Carbon ion radiotherapy is an emerging cancer treatment modality that has a greater therapeutic window than conventional photon radiotherapy. To maximize the efficacy of this extremely scarce medical resource, it is important to identify predictive biomarkers of higher carbon ion relative biological effectiveness (RBE) over photons. We addressed this issue by focusing on cellular antioxidant capacity and investigated 64Cu(II)-diacetyl-bis(N4-methylthiosemicarbazone) (64Cu-ATSM), a potential radioligand that reflects an over-reduced intracellular environment. We found that the carbon ion RBE correlated with 64Cu-ATSM uptake both in vitro and in vivo. High RBE/64Cu-ATSM cells showed greater steady-state levels of antioxidant proteins and increased capacity to scavenge reactive oxygen species in response to X-rays than low RBE/64Cu-ATSM counterparts; this upregulation of antioxidant systems was associated with downregulation of TCA cycle intermediates. Furthermore, inhibition of nuclear factor erythroid 2-related factor 2 (Nrf2) sensitized high RBE/64Cu-ATSM cells to X-rays, thereby reducing RBE values to levels comparable to those in low RBE/64Cu-ATSM cells. These data suggest that the cellular activity of Nrf2-driven antioxidant systems is a possible determinant of carbon ion RBE predictable by 64Cu-ATSM uptake. These new findings highlight the potential clinical utility of 64Cu-ATSM imaging to identify high RBE tumors that will benefit from carbon ion radiotherapy.

The growing importance of antitumour immunity by cancer immunotherapy has prompted studies on radiotherapy-induced immune response. Previous studies have indicated that programmed cell death-1 ligand (PD-L1) expression is regulated by DNA damage signalling. However, PD-L1 up-regulation after radiotherapy has not been fully investigated at the clinical level, particularly in the context of expression of DNA repair factors. The present study examined the correlation of mRNA expression between PD-L1 and non-homologous end joining (NHEJ) factors using The Cancer Genome Atlas database analysis. Among NHEJ factors, Ku80 mRNA expression was negatively correlated with PD-L1 mRNA expression levels in several types of cancer (colon adenocarcinoma, breast invasive carcinoma, skin cutaneous melanoma, lung adenocarcinoma, head and neck squamous cell carcinoma, uterine corpus endometrial carcinoma, cervical squamous cell carcinoma and endocervical adenocarcinoma). To verify the negative correlation in clinical samples, the present study analysed whether Ku80 expression levels affected PD-L1 up-regulation after radiotherapy using cervical squamous cell carcinoma samples. Quantitative evaluation using software analysis of immunohistochemically stained slides revealed that patients with low Ku80 positivity in biopsy specimens demonstrated increased PD-L1 expression levels after 10 Gy irradiation (Spearman's rank correlation coefficient=-0.274; P=0.017). Furthermore, PD-L1 induction levels in tumour cells after 10 Gy of irradiation were significantly inversely correlated with Ku80 expression levels (Spearman's rank correlation coefficient=-0.379; P<0.001). The present study also confirmed that short interfering RNA-mediated Ku80 depletion was associated with greater X-ray-induced PD-L1 up-regulation in HeLa cells. These results indicated that radiotherapy could enhance PD-L1 induction in tumour cells with low Ku80 expression in a clinical setting. Furthermore, these data highlighted Ku80 as a potential predictive biomarker for immune checkpoint therapy combined with radiotherapy.

Single-stranded DNA (ssDNA) arising as an intermediate of cellular processes on DNA is a potential vulnerability of the genome unless it is appropriately protected. Recent evidence suggests that R-loops, consisting of ssDNA and DNA-RNA hybrids, can form in the proximity of DNA double-strand breaks (DSBs) within transcriptionally active regions. However, how the vulnerability of ssDNA in R-loops is overcome during DSB repair remains unclear. Here, we identify RAP80 as a factor suppressing the vulnerability of ssDNA in R-loops, chromosome translocations, and deletions during DSB repair. Mechanistically, RAP80 prevents unscheduled nucleolytic processing of ssDNA in R-loops by CtIP. This mechanism promotes efficient DSB repair via transcription-associated end joining dependent on BRCA1, Polθ, and LIG1/3. Thus, RAP80 suppresses the vulnerability of R-loops during DSB repair, thereby precluding genomic abnormalities in a critical component of the genome caused by deleterious R-loop processing.