Detection of genome instability by 53BP1 expression as a long-lasting health effect in human epidermis surrounding radiation-induced skin cancers Authors: Katsuya Matsuda, Hirokazu Kurohama, Yutaka Kuwatsuka, Akira Iwanaga, Hiroyuki Murota, Masahiro Nakashima Published: 2024 ABSTRACT AbstractWe previously reported endogenous activation of the DNA damage response (DDR) in the epidermis surrounding basal cell carcinoma resected from Nagasaki atomic bomb survivors, suggesting the presence of genomic instability (GIN) in the survivors as a late effect of radiation. Dual-color immunofluorescence (IF) analysis of TP53-binding protein-1 (53BP1) and a proliferative indicator, Ki-67, to elucidate GIN in tumor tissues revealed that abnormal 53BP1 expression is closely associated with carcinogenesis in several organs. The present study aimed to confirm the presence of radiation-induced GIN in the non-neoplastic epidermis of patients with radiation-induced skin cancer. Formalin-fixed paraffin-embedded tissues were obtained from all participants between 2008 and 2019 at the Nagasaki University Hospital. 53BP1 nuclear expression was examined using dual-color IF analysis with Ki-67 expression to assess the extent and integrity of the DDR. Expressions of gamma-H2AX, p53 and p21 were also analyzed using the dual-color IF analysis for their association with 53BP1. The results of this study provide evidence for sporadic activation of the DDR in medically irradiated and ultraviolet-exposed epidermis as a long-lasting radiation effect, which is a predisposition to skin cancer. Furthermore, the incidence of abnormal 53BP1 expression in cancer cells was higher than in non-neoplastic epidermal cells surrounding cancer, suggesting a correlation between the type of 53BP1 and the malignant potential of skin tumors. This study highlights the usefulness of dual-color IF for 53BP1 (and Ki-67) as an indicator to estimate the level of GIN as a long-lasting health effect of radiation exposure.Keywords: radiation-induced cancer, skin, UV, 53BP1, DNA DSBs INTRODUCTION Seventy-eight years have elapsed since the two atomic bombs (A-bombs) exploded in Hiroshima and Nagasaki, Japan, in August 1945. Radiation etiology is suggested in the occurrence of basal cell carcinoma (BCC) of the skin in A-bomb survivors [1–3]. There is an excess relative risk per Gy of 0.48 for individuals receiving doses <1 Gy, whereas excess relative risk increases to 2.64 per Gy for those who absorb >1 Gy [4]. Therefore, the BCC incidence rate shows a dose–response relationship that decreases significantly with increasing distance from the hypocenter [4]. We have previously demonstrated an increased risk of developing BCC among Nagasaki A-bomb survivors who were exposed to radiation at a closer distance, providing evidence that A-bomb radiation is significantly associated with skin carcinogenesis [5]. Our previous report revealed endogenous activation of the DNA damage response (DDR) in the epidermis surrounding the BCC in the closer distance group, suggesting the presence of genomic instability (GIN) in the survivors as a late effect of A-bomb radiation, which may indicate a predisposition to cancer. In our previous studies, we used dual-color immunofluorescence (IF) analysis for TP53-binding protein-1 (53BP1) and the proliferative indicator Ki-67 to elucidate GIN in formalin-fixed paraffin-embedded (FFPE) tissues. 53BP1 is a nuclear DDR-associated molecule that rapidly localizes at DNA double-strand break (DSB) sites, forming nuclear foci (NF) to activate the downstream repair process [6–10]. The occurrence of DNA DSBs, which subsequently induce the DDR, is a hallmark of GIN [11]. Defective DDR can also lead to GIN, which is widely regarded as a key event in any carcinogenic process [12–15]. Abnormal 53BP1 expression has been reported to be closely associated with carcinogenesis in several organs [16–23]; for instance, large NF, such as those >1 μm, and diffuse patterns are significantly associated with high-grade urothelial carcinoma with chromosomal instability and poor prognosis in bladder cancers [19]. Additionally, abnormal 53BP1 expression has been associated with chromosomal instability, a prognostic parameter for gastric Mucosa associated lymphoid tissue (MALT) lymphoma [23]. Thus, we believe that analyzing 53BP1 expression using IF can be useful for estimating the level of GIN and malignant potential of human tumors. The present study was conducted to confirm the presence of radiation-induced GIN in the non-neoplastic epidermis using FFPE tissues of patients with radiation-induced skin cancer (RC). MATERIALS AND METHODS Participants Cases of RC are rare. Therefore, only three cases of RC, including one squamous cell carcinoma (SCC) and two Bowen’s disease (BD), which developed in the medically irradiated area decades after exposure, were available for this study (Fig. 1). As control individuals, randomly selected 23 cases of sporadic skin cancer (SC), including 12 cases from non-sun-exposed (SE) sites (five chests, three abdomens, two backs, one buttock and one thigh) and 11 cases from SE sites (four heads, five hands, one lower leg and one foot), were used in this study. Five cases of normal female skin surgically resected from the breast as a non-SE site under the diagnosis of breast cancer were also analyzed. To exclude age effects on 53BP1 expression in the epidermis, this study used cases from aged population as sporadic case, which could be comparable to age of radiation-induced cases. The clinicopathological profiles of these patients are summarized in Tables 1 and 2, and histological images of the non-neoplastic epidermis are depicted in Figure 2. FFPE tissues were obtained from all participants between 2008 and 2019 at the Nagasaki University Hospital. Macroscopic and histological (hematoxylin and eosin-stained) images of radiation-induced skin cancers. Case 1: SCC of the buttocks after radiotherapy for trichophytosis in a 73-year-old male atomic bomb survivor. Case 2: BD in the finger of an 80-year-old male radiologist. Case 3: BD of the dorsal foot after radiotherapy for trichophytosis in a 61-year-old female. Scale bars indicate 100 μm. Table 1Clinicopathologic profiles of patients with radiation-induced skin cancers AgeSexRadiationAge at radiationSitesHistotypeCase 1a73MTherapy-related X-ray for tinea capitis30ButtockSCCCase 2b80MOccupational X-ray>40 years since 24Right second fingerBDCase 361FTherapy-related X-ray for tinea capitis20Dorsal footBD aAtomic bomb survivor at 3.4 km from the hypocenter (estimated dose <1 mSv) when he was 3-year-old. bRadiologist engaged in X-ray fluoroscopy for long time, and suffered from radiation dermatitis in his right hand. Clinicopathologic profiles of patients with radiation-induced skin cancers aAtomic bomb survivor at 3.4 km from the hypocenter (estimated dose <1 mSv) when he was 3-year-old. bRadiologist engaged in X-ray fluoroscopy for long time, and suffered from radiation dermatitis in his right hand. Table 2Clinicopathologic profiles of control participants for radiation-induced skin cancersSubjects n Male/ femaleMedian age (range)Neoplasia (%)BDSCCNon-SE epidermis50/572.4(58–82)NANon-SE neoplastic1211/176.8(65–88)10(83.3)2(16.7)SE neoplastic1111/076.1(47–88)110NA = not applicable. Clinicopathologic profiles of control participants for radiation-induced skin cancers NA = not applicable. Histological images of non-neoplastic epidermis (hematoxylin and eosin-stained). (A) Nonexposed skin of the female breast. (B) Nonexposed skin surrounding skin cancer (SC). (C) Exposed skin surrounding SC. (D) Nonexposed and nonirradiated skin in Case 1. (E) Irradiated skin surrounding SC in Case 2. Scale bars indicate 100 μm. This study was performed retrospectively in accordance with the tenets of the Declaration of Helsinki. The Ethics Committee of Nagasaki university hospital approved the study (approval date: 20 November 2018; #18111921) and waived the need for informed consent. All methods were carried out in accordance with relevant guidelines and regulations. Patient profiles were anonymized by coding and collectively summarized with the obtained data as the final dataset. Patients were able to opt out of the study by following the instructions provided on the institute website. IF analysis for 53BP1, gamma-H2AX, p53 and p21 expression 53BP1 nuclear expression was examined using dual-color IF analysis with Ki-67 expression to assess the extent and integrity of the DDR. After deparaffinization and antigen retrieval by microwave treatment in Target Retrieval Solution (pH 6.0; Agilent Technologies, Santa Clara, CA, USA), tissue sections of 4 μm thickness were preincubated with Dako Protein Block, Serum-Free (DakoCytomation, Glostrup, Denmark). For dual-color IF, the sections were incubated with anti-53BP1 rabbit polyclonal antibody (1:1000; A200-272A; Bethyl Laboratories, Montgomery, TX) and anti-Ki-67 mouse antibody (1:50; MIB-1; DakoCytomation, Glostrup, Denmark) for 1 h at 20°C. Then, the slides were incubated with Alexa Fluor 488-conjugated goat anti-rabbit (Molecular Probes Inc., Eugene, OR, USA) and Alexa Fluor 594-conjugated goat anti-mouse (Molecular Probes Inc., Eugene, OR, USA) antibodies. All samples were counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (Vysis Inc., Downers Grove, IL, USA), analyzed and photographed using a high-standard all-in-one fluorescence microscope (Biorevo BZ-X710; KEYENCE Japan, Osaka, Japan) with the Z-stack function, accumulating images of ~30 slices. All signals for 53BP1 nuclear expression were analyzed in more than five viewing areas from the normal epidermis and cancer tissues in each case at 1000-fold magnification. The 53BP1 signals were measured using image analysis software provided with the Biorevo BZ-X710 microscope. According to our previous report [18], 53BP1 immunoreactivity can be classified into five types based on the number and size of NF: (i) stable type, no nuclear staining; (ii) low DDR type, one or two discrete NF; (iii) high DDR type, three or more discrete NF; (iv) diffuse type, intense heterogeneous nuclear staining and (v) large NF type, discrete NF >1.0 μm. In this study, Types (iii), (iv) and (v) were considered as abnormal 53BP1 expression. In addition, nuclei showing colocalization of 53BP1 NF and Ki-67 expression by double staining were considered DDR-abnormal cells. Because DDR is normally associated with cell cycle arrest, coexpression of 53BP1 NF and Ki-67 can be considered an indicator of an impaired DDR pathway. The percentage of cells expressing each protein type was calculated for each case. Representative images of 53BP1 expression are depicted in Figure 3. Types of TP53-binding protein 1 expression (green) by dual-color immunofluorescent analysis with Ki-67 expression (red). (A and F) Stable DDR type, no or faint nuclear staining. (B and G) Low DDR type, one or two discrete NF. (C and H) High-DDR type, three or more discrete NF. (D and I) Diffuse type, intense heterogeneous nuclear staining. (E and J) Large NF type, discrete NF >1.0 μm. The incidence of high DDR, large foci or diffuse types in the nuclei is considered as abnormal 53BP1 expression. Scale bars indicate 2 μm. Furthermore, expressions of gamma-H2AX (γH2AX) (anti-gamma-H2AX mouse monoclonal antibody (9F3) 1:1000; Abcam, Tokyo, Japan), p53 (anti-p53 mouse monoclonal antibody (clone Do-7) 1:100; DakoCytomation, Glostrup, Denmark) and p21 (anti-p21 mouse monoclonal antibody (clone DSC60) 1:100; Cell Signaling, MA, USA) were also analyzed using the same dual-color IF method for their association with 53BP1. The association between p21 and Ki-67 (anti-Ki67 rabbit monoclonal antibody (SP6) 1:200; Abcam, Tokyo, Japan) was also analyzed using dual-color IF. Statistical analyses The Student’s t-test was used to assess differences in the incidence of abnormal 53BP1 expression between the histological types of epidermal lesions. All statistical analyses were performed using StatView v5.0 (SAS Institute, Cary, NC, USA) and were two-tailed. A P-value of <0.05 was considered statistically significant. A homogeneity of variance among each group in both normal and cancer cells has been already confirmed by the Levene’s test, to apply the Student’s t-test in this analysis. RESULTS 53BP1 expression in non-neoplastic epidermis The results of IF analyses for 53BP1 nuclear expression in the non-neoplastic epidermis are shown in Table 3, and representative images are depicted in Figure 4. One case from 11 cases of sporadic SE subjects included too tiny amount of non-neoplastic epidermis to evaluate 53BP1 expression. Therefore, this study could analyze 11 cases of carcinoma but 10 cases of non-neoplastic epidermis as sporadic SE subjects. The incidence of abnormal 53BP1 expression in the non-neoplastic epidermis was low in the non-SE site of normal female breasts (0.2 ± 0.3%) but high in the non-SE site surrounding SC, SE site surrounding SC and irradiated site surrounding RC (5.4 ± 4.4%, 6.9 ± 4.1% and 12.3 ± 3.8%, respectively). Statistical analysis revealed significant differences between non-SE normal sites and non-SE sites surrounding SC (P = 0.0212) or SE sites surrounding SC (P = 0.0033) or irradiated sites surrounding RC (P = 0.0003) and between non-SE sites surrounding the SC and irradiated sites surrounding RC (Fig. 5). Although the statistical significance is not evident, the incidence of abnormal DDR type was higher in SE sites surrounding SC (4.3 ± 2.9%) and irradiated sites surrounding RC (3.6 ± 1.2%) than in non-SE sites (0.8 ± 1.0%). Table 3Incidence of types of 53BP1 expression in epidermis by IFParticipants n Male/ femaleCounted nucleiKi-67 LI (%)Type of 53BP1 expression (%)Abnormal DDR type (%) StableLow DDRHigh DDRLarge fociDiffuseAbnormalNon-neoplastic epidermisNon-SE50/517828.1 ± 2.790.2 ± 8.29.8 ± 7.90.1 ± 0.200.1 ± 0.20.2 ± 0.30.8 ± 1.0 Non-SE surrounding SC1211/148216.2 ± 2.577.8 ± 10.316.8 ± 6.44.1 ± 3.01.2 ± 2.30.02 ± 0.045.4 ± 4.41.6 ± 0.8 SE surrounding SC1010/057289.0 ± 4.667.9 ± 9.125.2 ± 6.16.8 ± 4.20.1 ± 0.306.9 ± 4.14.3 ± 2.9 Irradiated surrounding RC32/127398.4 ± 3.059.6 ± 6.928.0 ± 3.110.4 ± 4.21.8 ± 1.20.1 ± 0.112.3 ± 3.83.6 ± 1.2CarcinomasNon-SE1211/1568530.1 ± 10.665.4 ± 17.522.0 ± 10.09.1 ± 7.23.4 ± 4.60.8 ± 0.212.6 ± 9.211.4 ± 9.4 SE1110/1706229.0 ± 7.565.5 ± 19.721.4 ± 9.311.2 ± 9.01.9 ± 6.2013.1 ± 14.49.8 ± 6.1 RC32/1218714.3 ± 1.748.4 ± 17.028.5 ± 6.411.5 ± 8.49.5 ± 11.92.1 ± 2.323.1 ± 19.05.6 ± 2.3All values are presented as mean ± standard deviation.LI = labelling index.Abnormal type of 53BP1 expression includes high DDR, large foci and diffuse types.Abnormal DDR type: 53BP1 and Ki-67-double positive. Incidence of types of 53BP1 expression in epidermis by IF All values are presented as mean ± standard deviation. LI = labelling index. Abnormal type of 53BP1 expression includes high DDR, large foci and diffuse types. Abnormal DDR type: 53BP1 and Ki-67-double positive. Dual-labeled IF of TP53-binding protein 1 (green) and Ki-67 (red) expression in the non-neoplastic epidermis. (A) Nonexposed epidermis of the female breast. (B) Nonexposed epidermis surrounding SC. (C) Exposed epidermis surrounding SC. (D) Nonexposed and nonirradiated skin in Case 1. (E) Irradiated skin surrounding SC in Case 2. Comparison of median incidences of abnormal TP53-binding protein 1 expression in normal epidermal cells and skin cancer cells. * P < 0.05, ** P < 0.005 by Student’s t-test. ▲ indicates the mean value. 53BP1 expression in skin cancers The types of 53BP1 nuclear expression in skin cancers determined by IF analyses are summarized in Table 3, and representative images with Ki-67 expression are depicted in Figure 6. The incidence of abnormal 53BP1 expression in cancer cells of all three groups was higher than that in the respective non-neoplastic epidermis as background (12.6 ± 9.2% vs. 5.4 ± 4.4% in the non-SE site, 13.1 ± 14.4% vs. 6.9 ± 4.1% in the SE site, 23.1 ± 19.0% vs. 12.3 ± 3.8% in the irradiated site), although statistical significance was not evident between them. A comparison of cancer cells and epidermal cells in 22 sporadic cases showed increased incidence of abnormal 53BP1 expression in 18 cases (81.8%). (Supplementary Fig. 1). The incidence of abnormal 53BP1 expression was higher in RC than in SC, although this difference was not statistically significant. Furthermore, focusing on the large NF (LF) type and diffuse types, reported as indicators of chromosomal instability in our previous studies [18, 19], the incidence of these types in RC (11.6 ± 10.7%) was higher than that in SC at the non-SE (4.2 ± 4.8%, P = 0.0608) and SE (1.9% ± 6.2%, P = 0.0584) sites. The incidence of abnormal DDR type was higher in SC at the non-SE (11.4 ± 9.4%) and SE (9.8 ± 6.1%) sites than in RC (5.6 ± 2.3%); however, no significant differences were observed between them. In two cases of radiation-induced BD (Cases 2 and 3), the incidence of abnormal 53BP1 expression increased in the order of non-SE epidermis < irradiated epidermis < cancer cells (Supplementary Fig. 1). However, in the case of radiation-induced SCC from an A-bomb survivor (Case 1), although the incidence of abnormal 53BP1 expression was higher in the irradiated epidermis than in the non-SE epidermis, it was lower in cancer cells than in the irradiated epidermis (Supplementary Fig. 1). Dual-labeled IF of TP53-binding protein 1 (green) and Ki-67 (red)/TP53 (red) expressions in BDs at nonexposed site, exposed site and irradiated site in Case 2. Dual-color IF analyses Dual-color IF analysis of 53BP1 and Ki-67/TP53 revealed that 53BP1 NF rarely colocalized with Ki-67 signals, whereas 53BP1 signals, including large foci and abnormal type, frequently colocalized with TP53-block positivity as a type of mutant TP53 cancer cells of BD (Fig. 6). Double-labeled IF for 53BP1 and γH2AX (another DDR-related molecule) to detect DNA DSBs demonstrated frequent coexpression of both NF in cancer cells (Fig. 7). Furthermore, dual-color IF analysis for Ki-67 and p21 (an inhibitor of the cell cycle and apoptosis) showed that Ki-67 signals were found throughout the basal and top layers, whereas p21 signals were observed in the upper half of BD tissues (Fig. 8). Frequent colocalization of both signals (yellow) was evident in the upper part of the cancer cells at both the SE and irradiated sites (Fig. 8). Dual-labeled IF of TP53-binding protein 1 (green) and γH2AX (red) expression in cancer cells of patients with BD. Scale bars indicate 20 μm. 53BP1, TP53-binding protein 1. Dual-labeled IF of Ki-67 (red) and P21 (green) in BD at nonexposed site, exposed site and irradiated site from Case 2. Scale bars indicate 100 μm. DISCUSSION Radiation etiology in skin carcinogenesis, particularly BCC, is epidemiologically suggested not only among Japanese A-bomb survivors but also among patients treated with radiotherapy for benign skin lesions, such as tinea capitis, and workers with occupational exposure, including radiologists, radiologic technologists and Russian nuclear facilities (Mayak Production Association) [24–28]. An epidemiological study among A-bomb survivors also suggested that the incidence of BCC among individuals who received a high dose of radiation was significantly increased after 30 years, and a high risk of BCC persisted for a long time, such as >50 years after A-bombing, indicating a long-lasting radiation effect on the skin of A-bomb survivors [24]. This study detected radiation effects in the epidermis surrounding radiotherapy- and occupational exposure-associated radiation-induced skin cancers (RC, including BD and SCC) using dual-color IF analysis for 53BP1 expression as an indicator of DDR. Our previous study showed sporadic activation of the DDR in the epidermis surrounding BCC in A-bomb survivors who were heavily exposed to A-bomb radiation at a distance proximal to the hypocenter [5]. This study also revealed high levels of abnormal 53BP1 expression in the epidermis surrounding radiotherapy- and occupational exposure-associated RC, providing further evidence for sporadic activation of the DDR in the irradiated epidermis as a long-lasting radiation effect. Furthermore, the present study suggested that the SE epidermis surrounding sporadic SC as well as irradiated epidermis expressed a higher level of abnormal 53BP1 than the non-SE epidermis. Because non-SE epidermis from resected female breasts, which are naïve to ultraviolet (UV) exposure, expressed only 0.2% of abnormal 53BP1, continuous UV exposure could also induce sporadic activation of the DDR in the SE epidermis, as a predisposing factor for the development of skin cancers. However, non-SE epidermis surrounding SC in 92% of male patients expressed a similar level of abnormal 53BP1 expression with respect to SE epidermis surrounding SC, suggesting the involvement of genotoxic factors other than UV and ionizing radiation in skin carcinogenesis at non-SE sites. Similarly, our previous study revealed an increased number of 53BP1 NF in the normal urothelium surrounding bladder carcinoma [19]. Our previous study identified several 53BP1 NF in human skin tumor tissues, including benign lesions [17]. Actinic keratosis, a precancerous lesion, showed a high level of DDR-type 53BP1 immunoreactivity and BD as carcinoma in situ. Furthermore, BCC exhibited a high level of abnormal 53BP1 expression with a mutant type of nuclear accumulation of TP53 immunoreactivity, suggesting disruption of the DDR [17]. This study also showed an increase in abnormal 53BP1 expression in cancer cells compared with the respective background epidermal cells in sporadic cases. In addition, BD exhibited frequent colocalization of abnormal 53BP1 and nuclear TP53 accumulation. Among the three cases of RC, two BD cases showed a stepwise increase in abnormal 53BP1 expression in RC carcinogenesis in the order of non-SE epidermis < irradiated epidermis < cancer, whereas one SCC case showed an increase in abnormal 53BP1 expression from non-SE epidermis to irradiated epidermis but a decrease in cancer. We have previously demonstrated a correlation between the expression of 53BP1 and the malignant potential of human tumors, including uterine cervical [18] and esophageal cancers [21]. Our previous observations showed that the number of cells expressing 53BP1 NF was significantly low in the deeply invasive parts of SCC lesions [18, 21]. We speculate that tumor cells exhibiting invasive patterns severely abrogate their ability to respond to DNA damage, possibly indicating an advanced phase of carcinogenesis. In this study, we observed frequent colocalization of 53BP1 and γH2AX NF in skin cancer cells. Our previous experiments based on IF analysis of FFPE tissues showed that the presence of 53BP1 NF was increased in irradiated rat thyroid glands in a dose-dependent manner and that 53BP1 and γH2AX NF were frequently colocalized in human follicular thyroid cancers as well as irradiated rat thyroid glands, suggesting endogenous activation of the DDR pathways in tumor cells as a hallmark of GIN [29]. In contrast, another dual-color IF demonstrated the colocalization of p21 and Ki-67 in the upper part of BD lesions in SE and irradiated sites but not in non-SE sites. Because p21, a cyclin-dependent kinase inhibitor, promotes growth arrest and cellular senescence via the ATM–TP53–P21 pathway after irradiation, aberrations in these molecules can induce solid cancers owing to the failure of growth arrest and cellular senescence [30, 31]. Therefore, the colocalization of p21 and Ki-67 in BD lesions at SE and irradiated sites may indicate an aberration of this pathway during skin carcinogenesis induced by chronic UV exposure or ionizing radiation. This study provides evidence for sporadic activation of the DDR in medically irradiated and UV-exposed epidermis as a long-lasting radiation effect that predisposes the skin to cancer. Furthermore, the incidence of abnormal 53BP1 expression in cancer cells was high in non-neoplastic epidermal cells surrounding cancer in almost all cases, suggesting a correlation between the type of 53BP1 and the malignant potential of skin tumors. A series of double-labeled IF studies comparing 53BP1, Ki-67, TP53, γH2AX and p21 expressions may be useful to histologically estimate the status of the DDR pathway during radiation-induced skin carcinogenesis. As a limitation of our analysis, this study demonstrated the difference of type of 53BP1 expression between radiation-associated and sporadic groups; however, the statistical significance was not proven evident in some comparisons such as in cancer tissues. It may be due to small number of subjects because of the scarcity of medical radiation-induced skin cancers. Further future studies are required to confirm results in the present study. CONFLICT OF INTEREST The authors declare no conflict of interest. FUNDING This work was funded by the Atomic Bomb Disease Institute Nagasaki University, and the Program of the Network-Type Joint Usage/Research Center for Radiation Disaster Medical Science. Supplementary Material Presentation at a conference: The 5th International Symposium of the Network-type Joint Usage/Research Center for Radiation Disaster Medical Science (8 February 2021) Contributor Information Katsuya Matsuda, Department of Tumor and Diagnostic Pathology, Atomic Bomb Disease Institute, Nagasaki University Graduate School of Biomedical Sciences, 1-12-4 Sakamoto, Nagasaki 852-8523, Japan. Hirokazu Kurohama, Department of Tumor and Diagnostic Pathology, Atomic Bomb Disease Institute, Nagasaki University Graduate School of Biomedical Sciences, 1-12-4 Sakamoto, Nagasaki 852-8523, Japan. Yutaka Kuwatsuka, Department of Dermatology, Nagasaki University Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan. Akira Iwanaga, Department of Dermatology, Nagasaki University Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan. Hiroyuki Murota, Department of Dermatology, Nagasaki University Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan. Masahiro Nakashima, Department of Tumor and Diagnostic Pathology, Atomic Bomb Disease Institute, Nagasaki University Graduate School of Biomedical Sciences, 1-12-4 Sakamoto, Nagasaki 852-8523, Japan. REFERENCES Associated Data Supplementary Materials