- Email: [email protected]
Physical, Heritable and Age-Related Factors as Modifiers of Radiation Cancer Risk in Patched Heterozygous Mice
Int. J. Radiation Oncology Biol. Phys., Vol. 73, No. 4, pp. 1203–1210, 2009 Copyright Ó 2009 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/09/$–see front matter
doi:10.1016/j.ijrobp.2008.10.068
BIOLOGY CONTRIBUTION
PHYSICAL, HERITABLE AND AGE-RELATED FACTORS AS MODIFIERS OF RADIATION CANCER RISK IN PATCHED HETEROZYGOUS MICE SIMONETTA PAZZAGLIA, PH.D.,* EMANUELA PASQUALI, M.SC.,y MIRELLA TANORI, PH.D.,* MARIATERESA MANCUSO, PH.D.,* SIMONA LEONARDI, PH.D.,* VINCENZO DI MAJO, PH.D.,* SIMONETTA REBESSI, PH.D.,* AND ANNA SARAN, PH.D.* * Section of Toxicology and Biomedical Sciences, Biotechnologies, Agro-Industry and Health Protection Department, ENEA CR Casaccia, 00123 Rome, Italy, and y Department of Experimental Oncology, Istituto Nazionale Tumori, 20133 Milan, Italy Purpose: To address the tumorigenic potential of exposure to low/intermediate doses of ionizing radiation and to identify biological factors influencing tumor response in a mouse model highly susceptible to radiogenic cancer. Methods and Materials: Newborn Ptc1 heterozygous mice were exposed to X-ray doses of 100, 250, and 500 mGy, and tumor development was monitored for their lifetime. Additional groups were irradiated with the same doses and sacrificed at fixed times for determination of short-term endpoints, such as apoptosis and early preneoplastic lesions in cerebellum. Finally, groups of Ptc1 heterozygous mice were bred on the C57BL/6 background to study the influence of common variant genes on radiation response. Results: We have identified a significant effect of low-intermediate doses of radiation (250 and 500 mGy) in shortening mean survival and inducing early and more progressed stages of tumor development in the cerebellum of Ptc1+/– mice. In addition, we show that age at exposure and heritable factors are potent modifiers of radiationrelated cancer risk. Conclusions: The Ptc1 knockout mouse model offers a highly sensitive system that may potentially help to improve understanding and quantification of risk at low doses, such as doses experienced in occupational and medical exposures, and clarify the complex interactions between genetic and environmental factors underlying cancer susceptibility. Ó 2009 Elsevier Inc. Patched1, Medulloblastoma, Cancer risk, Genetic background, Ionizing radiation.
Animal studies play an important role in improving our understanding of radiation carcinogenesis (1, 2). The use of tumor data from animal studies is necessary as a complement of epidemiological studies of human populations to develop estimates of radiation cancer risk at low doses. In addition, animal experiments provide valuable insights into the mechanisms of radiation interaction with living cells and organisms, allowing clarification of the pathways of tumorigenesis and of the factors modifying radiation risks. Animal studies, however, are hampered by the requirement of very large animal numbers to reach statistical significance, particularly at low radiation doses. Thus, genetically manipulated, radiation-susceptible mouse models represent a powerful tool to help assessment of risk. In addition, a strong need exists for systems that provide information on the mechanisms whereby a ‘‘hit’’ normal cell develops into a tumor.
The Ptc1 knockout mouse model is a well-established model of carcinogenesis, particularly in the brain, skin, and soft tissues (3–5). Radiation is known to elevate cerebellum tumor incidence and to promote basal cell carcinoma (BCC) precursor lesions to progress to large infiltrative BCC (6), but little is known of low-dose radiation effects or the shape of the dose–response for radiation-induced cancers at low doses. Understanding of other factors that are known to modify radiation-related tumorigenesis is also incomplete. To investigate cancer susceptibility to low to intermediate doses, we carried out experiments for assessment of tumor responses to 100, 250, and 500 mGy of X-rays in Ptc1+/– mice. Dose dependence was also established for several biological endpoints such as tumor-free survival, apoptotic cell death, and early-preneoplastic areas in cerebellum. Age at irradiation and genetic background were also examined as modifiers of radiation-related cancer risk in cerebellum.
Reprint requests to: Simonetta Pazzaglia, Ph.D., Section of Toxicology and Biomedical Sciences, Biotechnologies, Agro-Industry and Health Protection Department, ENEA CR Casaccia, 00123 Rome, Italy. Tel: (+39) 06-3048-6535; Fax: (+39) 06-3048-3644; E-mail: [email protected]
Conflict of interest: none Acknowledgments—This work was supported by EU Contract FI6RCT-2003-508842 RISC-RAD. Received Aug 5, 2008, and in revised form Oct 15, 2008. Accepted for publication Oct 16, 2008.
INTRODUCTION
1203
1204
I. J. Radiation Oncology d Biology d Physics
Analysis of Ptc1 LOH
METHODS AND MATERIALS Mice neo67/+
Volume 73, Number 4, 2009
+/–
Mice lacking one Ptc1 allele (Ptc , termed Ptc1 throughout the text), generated through disruption of exons 6 and 7 in 129S2 ES cells (4) and maintained on a CD1 or C57BL/6 background, were used in this study. CD1 is an outbred stock, and brother-sister mating was avoided to minimize consanguinity. C57BL/6-Ptc1+/– mice were generated by breeding the 129S2 mutation for at least 10 generations onto the C57BL/6NCrlBR (B6) mice purchased from Charles River Laboratorie (Sulzfeld, Germany) (7). Mice were genotyped using polymerase chain reaction primers specific to the neo insert and wildtype regions as described previously (4). Animals were housed under conventional conditions with food and water available ad libitum and a 12-h light-dark cycle. Experimental protocols were reviewed by the Internal Institutional Animal Care and Use Committee.
Irradiation For tumor induction, groups of Ptc1+/– mice on CD1 background were whole-body irradiated at postnatal Day 1 (P1) with 100 (n = 45), 250 (n = 45), or 500 mGy (n = 39), or left untreated (n = 51). For study of genetic background effects, Ptc1+/– mice on C57BL/ 6 background were whole-body irradiated at P1 with 3 Gy of X-rays (n = 15) or left unirradiated (n = 22). For age-at-irradiation effects, data from previous experiments with X-rays (3 Gy) delivered at P1, P2, P4, and P10 were reanalyzed (8–10). Irradiation was performed using a Gilardoni CHF 320 G X-ray generator (Gilardoni, Mandello del Lario, Lecco, Italy) operated at 250 kVp, 1 mA for 100 mGy (dose rate: 59 mGy/min), 5 mA for 250 and 500 mGy (dose rate: 327 mGy/min), 15 mA for 3Gy (dose rate: 1,000 mGy/ min), with filters of 2.0 mm Al and 0.5 mm Cu (HVL=1.6 mm Cu).
Tumor quantification and histological analysis Mice were observed daily for their whole life span. Upon decline of health (i.e., severe weight loss, paralysis, ruffling of fur, or inactivity), they were euthanized and necropsied. Whole brain, dorsal skin, and any visible mass were fixed in 4% buffered formalin. Samples were then embedded in paraffin wax according to standard techniques. Cerebellar tumors and normal tissue were snap frozen. Tumor incidence was expressed as the percentage of mice with one or more tumors. Incidence of microscopic cerebellar lesions was determined on asymptomatic Ptc1+/– mice irradiated with 100 (n = 17), 250 (n = 12), or 500 mGy (n = 14) X-rays at P1, or left untreated (n = 16), and sacrificed at 8 weeks postirradiation. In total, 18 sections were examined for each cerebellum with an interval of 70 mm.
Radiosensitivity analysis Brains (n = 3 per time point) of irradiated and unirradiated Ptc1+/– pups were collected at 6 h postirradiation and fixed in 4% buffered formalin. Serial sections of cerebellar tissues were cut at 4-mm thickness and stained with hematoxylin and eosin. Digital images from midsagittal cerebellar sections of P1 pups, covering the entire cerebellar surface, were collected by IAS image-processing software (Delta Sistemi, Rome, Italy). Cells showing signs of nuclear chromatin condensation and morphologically normal cells in the external granule layer (EGL) of the cerebellum were counted using a double-blind method. Apoptotic values were calculated as the percentage of pyknotic nuclei relative to the total number of cells. The total number of cells examined ranged from 0.9 104 to 1.4 104.
Genomic and tumor DNA was extracted from spleens and medulloblastomas by Wizard SV Genomic DNA Purification System (Promega, Madison, WI). LOH analysis at the Ptc1 locus was performed by amplification and sequencing of exon 23 of the murine Ptc1 gene, which holds a polymorphic site at position 4016 between the targeted Ptc1 allele in the 129S2 strain and the wildtype allele from the CD1 stock. The primers used for the study were primer pair 8F/9R (11). Sequencing reactions were performed by means of dye terminator chemistry using a Big Dye Terminator version 3.1 Sequencing Cycle Kit and a 3730 DNA Analyzer (Applied Biosystems, Foster City, CA). Loss of polymorphic sites in medulloblastoma compared with normal tissue was considered LOH at the Ptc1 locus.
Statistics Survival results are reported as means SE; Mann-Whitney U test was used for determination of statistical differences between groups. Kaplan-Meier medulloblastoma-free survival curves were compared, and log-rank test p values were calculated. Analyses were done using GraphPad Prism version 4.02 for Windows (GraphPad Software, San Diego, CA). A p value of < 0.05 was considered statistically significant.
RESULTS Survival and pathology of Ptc1+/– mice after low-dose X-ray irradiation The tumor-free survival of Ptc1+/– mice exposed to 0, 100, 250, and 500 mGy of X-rays at P1 is shown in Fig. 1a. For comparison, survival of mice irradiated with 3 Gy is reported from previous work (7). Survival data are summarized in Fig. 1b. The mean survival was significantly reduced in Ptc1+/– mice exposed to 250 mGy (46.4 4.6; p = 0.0451) and 500 mGy (37.4 3.81; p = 0.0007) compared with unirradiated mice (56.0 3.7). Reduction in survival was also observed in Ptc1+/– mice irradiated with the lowest 100 mGy dose (48.8 3.3), although this was not statistically significant. The tumor spectrum of the different Ptc1+/– groups is shown in Fig. 2b. Unexposed mice showed the expected rates of spontaneously occurring Ptc1-associated tumors, i.e., medulloblastomas and soft tissue sarcomas. In irradiated groups, medulloblastomas occurred at an enhanced rate with increasing dose level and represented a competing cause of death relative to other tumors. There was a fourfold increase in medulloblastoma incidence in mice irradiated with 250 mGy (p = 0.0113), and a sixfold increase at 500 mGy (p < 0.0001) relative to controls. In mice exposed to 100 mGy, we found a nearly twofold increase over controls; however, this difference was not statistically significant, and a larger group size (i.e., n > 200) would be required. Although noninducible sarcomas remained the predominant tumor type in Ptc1+/– mice exposed to 100 and 250 mGy (47.6% and 35.7%), medulloblastoma was the most frequent tumor (50%) at 500 mGy. Tumors started between 8 and 12 weeks of age in all groups (Fig. 2a). Compared with previous experiments at 3 Gy (Fig. 2a, dashed line; median survival = 11 weeks) (8),
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
(a) 100 100 mGy 250 mGy
80
500 mGy 3000 mGy
60 40 20
Medulloblastoma free mice (%)
(a) 0 mGy
Percent survival
1205
100
0 mGy 100 mGy
80
250 mGy 60
500 mGy 3000 mGy
40 20 0 0
10
20
30
40 80 100
Time (weeks)
0 0
25
50
75
100
(b)
Weeks of age
(b) Dose (mGy)
ap
Number of mice
Mean survival (weeks ± SE)
0
51
100
45
56.2 ± 3.7 48.8 ± 3.3
250
45
46.4 ± 4.6a
500
39
37.4 ± 3.8b
= 0.0451; b p = 0.0007 in the Mann-Whitney U test
Fig. 1. Tumor-free survival of Ptc1+/– mice. (a) Survival of Ptc1+/– mice exposed to 0, 100, 250, and 500 mGy of X-rays at P1. Survival of mice irradiated with 3 Gy is shown for comparison (dashed line) (8). (b) Mean survival after exposure to X-rays (100, 250, or 500 mGy) compared with control Ptc1+/– mice.
the time course of medulloblastoma development was significantly delayed at 500 mGy (median survival = 39 weeks; p < 0.0001, log-rank test). Whereas 50% of the tumors (18/36) from the current experiments developed in mice older than 20 weeks, all medulloblastomas (17/17) from the 3-Gy experiment arose within 20 weeks postirradiation, and 60% of the animals had developed tumors by week 13. These results indicate that the kinetics of medulloblastoma development can be significantly influenced depending on radiation dose. The skin represents an additional target of ionizing radiation in the Ptc1+/– model because exposure to 3 Gy results in significant induction of BCC (6). Notably, Ptc1+/– mice irradiated with 100–500 mGy and observed for their lifetime did not develop skin tumors, revealing substantial differences in radiosensitivity between skin and cerebellum.
Quantitative dose–response relationships In Ptc1 heterozygous mice, we examined the dose– response curves in the low-dose range for several biological endpoints, including tumor-free survival, apoptotic cell death in the EGL of the cerebellum, early preneoplastic cerebellar lesions, and full medulloblastoma development. Tumor-free survival. The survival of irradiated Ptc1+/– mice in the dose range of 0–500 mGy is shown in Fig. 3. Data are well fitted using a nonthreshold linear dose dependence (r2 = 0.87). When the larger 0–3,000 mGy dose range was considered, a linear-quadratic dose dependence represented the best fit (r2 = 0.98; Fig. 3, inset).
Tumor type (%)
0 mGy
100 mGy
250 mGy 500 mGy
Medulloblastoma BCC Lymphoma Sarcoma Others
4 (7.8) 2 (3.9) 19 (37.3) 6 (11.8)
6 (14.3) 3 (7.1) 20 (47.6) 5 (11.9)
12 (28.6)a 18 (50.0)b 1 (2.4) 1 (2.8) 15 (35.7) 8 (22.2) 4 (9.52) 2 (5.5)
No of mice with tumors
31 (60.8)
31 (73.8)
29 (69.0) 28 (77.8)
51
42
No of autopsied mice
42
36
a p = 0.0113; b p<0.0001 in the log rank test
Fig. 2. Rate of medulloblastoma development after low-dose irradiation in Ptc1+/– mice. (a) Kaplan-Meier kinetic analysis of medulloblastoma in Ptc1+/– mice exposed to 100, 250, and 500 mGy of X-rays. Accelerated medulloblastoma rate in mice irradiated with 3 Gy is shown for comparison (dashed line; ref. 8). (b) Tumor spectrum in irradiated (100, 250, or 500 mGy) and control Ptc1+/– mice.
Radiation-induced cell death in granule cell progenitors (GCPs). GCPs are considered the cells of origin of medulloblastoma (12). At P1, proliferating GCPs are clustered over the surface of the developing cerebellum to form the EGL. By quantification of pyknotic nuclei in the EGL of P1-cerebella, we examined the apoptotic levels at 6 h after exposure to 100, 250, and 500 mGy. Significant differences
Fig. 3. Dose–effect relationship for mean survival time in Ptc1+/– mice exposed to low X-ray doses at P1. The fit was linear in the dose range 0–500 mGy. (Inset) Data were best fitted by a linear-quadratic dose dependence in the 0–3,000 mGy dose interval. Data points are means standard error.
1206
I. J. Radiation Oncology d Biology d Physics
Volume 73, Number 4, 2009
were statistically significant between Ptc1+/– mice irradiated with 250 (p = 0.0242) or 500 mGy (p = 0.0051) and unirradiated mice. The increase was not significant in mice irradiated with 100 mGy. We fit these data using a linear regression (r2 = 0.70; slope = 0.1). To test whether other models would provide a better fit, we tested a linear-quadratic model. This improved fit (r2 = 0.9). However, when only the 100 and 250 mGy data points were considered, the data showed a significant linear trend (r2 = 0.88). Data in the 0–3,000 mGy dose interval were best fitted using a linear-quadratic dose dependence (r2 = 0.90; Fig. 5C, inset). Analysis of medulloblastoma development in cerebellum of Ptc1+/– mice showed clear dose dependence at doses below 500 mGy (Fig. 5D and 5E). A very good linearity (r2 = 0.99) and a similar slope (i.e., 0.086) compared with the early-lesion endpoint (slope ratio = 1.2) suggest similar effectiveness of radiation for induction of preneoplastic and fully malignant growth. Data relative to the 0–3,000 mGy range were best fitted using a linear-quadratic dose dependence (r2 = 0.99; Fig. 5F, inset).
+/–
Fig. 4. Apoptotic response in the cerebellum of Ptc1 mice exposed to low X-ray doses at P1. (A) Hematoxylin and eosin stained sections of cerebellum at 6 h after irradiation with 100 mGy, (B) 250 mGy, (C) 500 mGy, and (D) unirradiated cerebellum. (E) Dose response for apoptosis in the external granule layer (EGL) of Ptc1+/– mice after irradiation with 100, 250, and 500 mGy of X-rays. Data in the interval 0–500 mGy were fitted with a linear regression. The best fit of the data in the 0–3,000 mGy dose interval is represented by a linear-quadratic dose dependence (inset). Data points represent the percentage of apoptotic cells out of the total number of cells in the EGL standard error.
in the number of cells undergoing apoptosis in irradiated compared with control cerebellum were detected at all doses (Fig. 4 A–D; p < 0.0001). Incidences were 2.1, 4.1, and 10.7% of GCPs at 100, 250, and 500 mGy, respectively. A linear regression yielded a good fit for the kinetics of cell inactivation over this range (r2 = 0.98; Fig. 4E). The best data fit was again represented by a linear-quadratic dose dependence (r2 = 0.99) when data in the 0–3,000 mGy range were analyzed (Fig. 4E, inset). Dose responses for early preneoplastic EGL lesions and medulloblastoma induction. Young Ptc1+/– mice develop a preneoplastic phenotype represented by ectopic EGL regions in the cerebellum of asymptomatic mice (8). We have shown previously that exposure to 3 Gy of X-rays substantially increases the size and frequency of such lesions (13), but the effects of low to intermediate doses have not been assessed. We performed histological examination of cerebella at 8 weeks after neonatal irradiation with 100, 250, and 500 mGy (Fig. 5A and 5B) and show that the frequency of cerebellar abnormalities in asymptomatic mice increases with radiation dose in the 0–500 mGy dose range (Fig. 5C). Differences in the frequency of cerebellar lesions
Ptc1 loss of heterozygosity (LOH) in medulloblastomas induced by low-dose exposure Biallelic Ptc1 loss represents an early and critical event in medulloblastoma development in Ptc1+/– mice exposed to radiation and the biological switch to malignancy in early cerebellar lesions (13). We asked whether the processes that drive carcinogenesis due to high dose exposures also contribute to low-dose radiation carcinogenesis. To investigate whether radiation-induced DNA damage and tumorigenic mechanisms differ at low doses, we assayed loss of the wildtype Ptc1 allele in medulloblastomas induced at doses below 500 mGy. Sequence analysis of DNA showed that, similar to tumors induced at higher dose, lack of the wildtype Ptc1 allele was found in all (9/9) medulloblastomas analyzed (Fig. 5 E). Age-at-irradiation effects We have previously shown that ionizing radiation induces medulloblastoma only in neonatal Ptc1+/– mice (9). The age– response relationship for tumor induction in the cerebellum after exposure to 3 Gy of X-rays at P1, P2, P4, and P10 is shown in Fig. 6a. There was a drastic decrease in tumor incidence with increasing age at irradiation from P1 to P10, age at which tumorigenic effects of radiation decreased out despite dramatically increasing size of cerebellum, and number of target cells. Sagittal sections of cerebella at P1, P2, P4, and P10 are shown in Fig. 6 b–e. The remarkable decrease in sensitivity with increasing age at irradiation suggests that developmental changes occurring after birth diminish sensitivity to radiation-induced malignancy in the central nervous system (CNS). Genetic factors In Ptc1 mice, genetic background strongly affects the spectrum of neoplasms (4, 7, 14). To test whether modifiers of
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
1207
Fig. 5. Tumorigenesis by low radiation doses in the cerebellum of Ptc1+/– mice. (A and B) Histological features of early cerebellar lesions (arrows) in Ptc1+/– mice irradiated at P1. (C) Dose–effect relationship for the induction of early cerebellar lesions in Ptc1+/– mice irradiated with 100, 250, and 500 mGy of X-rays; data were fitted by a linear dose-dependence (r2 = 0.7). (Inset) Best fit of the data in the 0–3,000 mGy dose interval. (D) Gross appearance of a normal brain from Ptc1 mice and relative DNA sequencing showing the presence of a polymorphism (T/C) at position 4016 of the Ptc1 gene; this site differs between the 129S2 strain (T allele), on which the Ptc1 knock out was engineered, and the CD-1 stock which is the source of the wildtype Ptc1 (C allele). (E) Brain with a large medulloblastoma (arrow) and electropherogram of tumor DNA showing only the targeted allele and demonstrating loss of the wildtype Ptc1 allele. (F) Dose response for full medulloblastoma development after irradiation with 100, 250, and 500 mGy of X-rays. (Inset) A linear quadratic dose dependence represents the best fit in the 0–3,000 mGy dose interval.
cancer susceptibility also modify radiation-induced cancer, neonatal Ptc1+/– mice on C57BL/6 background were exposed to 3 Gy of X-rays. We observed a spontaneous sevenfold higher incidence of medulloblastoma (40%) in C57BL/ 6-Ptc1+/– compared with CD1-Ptc1+/– mice (7.7%; p = 0.001, chi-square test; Fig. 7a and 7b). Notably, no induction of medulloblastoma over the spontaneous rate was observed in C57BL/6-Ptc1+/– mice after irradiation with 3 Gy (Fig. 7b). For comparison, the tumor response in neonatally irradiated CD1-Ptc1+/– mice is reported from a previous publication (8) (Fig. 7a). These findings suggest that gene–environment interactions are important determinants in cancer risks from radiation.
DISCUSSION Cancer risk caused by exposure to low doses of ionizing radiations is particularly important for assessing the safety of clinical and industrial applications of radiation. In this study, a mouse model of spontaneous and radiogenic cancer has been used to investigate the effects of exposure to doses below 500 mGy on tumor yields. We have been able to detect significant increases in brain tumor incidence at doses of 250 and 500 mGy of X-rays. In contrast, the skin, an additional radiation target in the Ptc1+/– model, was not susceptible to radiation doses below 500 mGy. Thus, although Ptc1-associated BCC and medulloblastoma show common features (i.e.,
I. J. Radiation Oncology d Biology d Physics
1208
Volume 73, Number 4, 2009
The experiments described here were designed to test whether the dose–response relationships for several parameters (i.e., life shortening, induction of apoptosis, and cancer induction) could be determined by limited-size animal experiments and to assess the shape of the dose–response curve. Experimental data from this study provide clear quantitative evidence that a linear dose dependence is adequate to describe responses for several biological endpoints in the dose range 0–500 mGy. By contrast, a linear-quadratic dependence represented the best fit for all endpoints when experimental data in the dose range 0–3,000 mGy were considered. Similar to previous experiments at 3 Gy, the same molecular event (i.e., specific loss of the wildtype Ptc1 allele) was detected in medulloblastomas induced by low radiation doses, showing that radiation is acting by inducing gene loss events. However, medulloblastoma development was significantly delayed at doses # 500 mGy, suggesting that radiation dose influences the time frame for Ptc1 LOH. For radiological protection, it is important to gain a better view of the common factors influencing radiation response, and the age at irradiation is known to influence cancer risk (16). Data on the effect of postnatal age at irradiation from follow-up studies of atomic bomb survivors show that relative cancer risks are higher at younger ages for several cancer types (17, 18). Our results provide quantitative evidence that neonatal Ptc1+/– mice are at greatest risk of radiation-associated brain cancer, possibly reflecting a larger proportion of dividing cells that could acquire tumor-initiating mutations in early postnatal cerebellum (19). Strong similarities exist between brain developmental processes in mice and humans. In mice, the formation of cerebellum spans embryonic and postnatal development, initiating at 9 days postcoitus and completing by postnatal Day 20. In humans the nervous system begins to form during the first trimester of gestation and protracts until after birth with extensive postnatal neurogenesis occurring in the cerebellum for a further 1 year. Compared with mouse, one might expect increased radiation
Fig. 6. Age-at-exposure effects for medulloblastoma (Mb) tumorigenesis after irradiation of Ptc1+/– mice with 3 Gy of X-rays. (a) Decrease in tumor incidence with increasing age at irradiation from P1 to P10. (b–e) Hematoxylin and eosin stained sections of mouse cerebellum at P1 (b), P2 (c), P4 (d), and P10 (e), showing progressive increase in the size during early postnatal development.
both are radiogenic tumors developing through early preneoplastic lesions that require biallelic Ptc1 loss for progression to full malignancy) (6, 13), skin and cerebellum show markedly different radiosensitivity to tumor induction. The basic concept that individual tissues differ in quantitative response to radiation also emerges in a study on cancer-prone TP53 heterozygous mice in which a single exposure to 100 mGy of X-rays decreased risk of lymphoma by extending tumor latency but increased risk of osteosarcoma (15). Because most environmental exposures occur at low doses, one of the main issues in radioprotection is to assess the shape of the dose–response curve in the low-dose region.
(a)
(b)
Medulloblastoma incidence( )
3 Gy (n=21)
100
CD1 0 Gy (n=51)
3 Gy (n=15)
100
C57BL/6 0 Gy (n=22)
75
75
50
50
25
25 0
0 0
10
20
30
Age (weeks)
40
50
0
10
20
30
40
50
Age (weeks)
Fig. 7. Influence of genetic background on development of spontaneous and radiogenic medulloblastoma in Ptc1+/– mice. (a) Neonatal irradiation of CD1-Ptc1+/– mice with 3 Gy of X-rays induced a 10-fold increase in medulloblastoma final rate relative to controls (8). (b) Irradiation of neonatal C57BL/6-Ptc1+/– mice with the same dose did not enhance medulloblastoma. A sevenfold higher incidence of spontaneous medulloblastoma was observed in C57BL/6-Ptc1+/– mice compared with CD1-Ptc1+/– mice (p = 0.001, chi-square test).
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
risk of medulloblastoma for children younger than 1 year. Importantly, exposure to radiation therapy has been recognized as the most important risk factor for the development of a new CNS tumor in survivors of childhood cancers (19–23). In particular, excess relative risk is highest among children exposed at younger than 5 years, suggesting a greater susceptibility of the developing brain to radiation (24). Our tumor data in mice confirm that the age at exposure is a consistent modifier of radiation-related CNS cancer risk and that significant differences exist between the developing and mature nervous system in tumor response to exogenous DNA damage. Currently, within age and sex groups, radiation risk is assumed to be uniformly distributed across populations. However, genetic variations may cause individual variability in biological responses to radiation and increased risk in susceptible individuals (25). In the presence of high penetrance alleles, genetics plays a major role in determining individual risk. The Ptc1 gene, indeed, represents a paradigm of radiation hypersensitivity, leading to cancer. However, frequencies of such high penetrance alleles in humans are low, and greater contribution to radiation risks is probably due to genes that are less penetrant but more frequent. We show here that the genetic background hosting the Ptc1 mutation can dramatically alter the individual risk for developing a specific tumor. In fact, whereas spontaneous medulloblastoma was significantly increased when the Ptc1+/– mutation was expressed in the C57BL/6 background, C57BL/6-Ptc+/–
1209
mice were completely resistant to medulloblastoma induction by radiation, suggesting that different sets of genes may control susceptibility to spontaneous and radiation-induced cancer. These findings indicate that complex interactions between genetic and environmental factors underlie susceptibility to radiogenic cancer in the CNS. Interestingly, familial aggregation for development of radiation-associated meningioma has been recently shown, supporting the hypothesis of genetic predisposition for radiogenic CNS tumors in humans (26). Thus, a better view of the common variant genes that influence radiation response would contribute to the development of more refined approaches to assessment of radiation health risk in humans. In summary, we have employed a mouse model of radiogenic cancer to assess the tumorigenic potential of exposure to low to intermediate doses of ionizing radiation and to identify biological factors influencing tumor response. Our in vivo findings identify a linear dose–response relationship in the low-dose region for several biological endpoints, such as tumor-free survival, apoptosis, and tumor induction. We also provide evidence that age at exposure and heritable factors are potent modifiers of radiation-related cancer risk in a susceptible mouse model. Although a genetically susceptible population represents a worst-case scenario, the use of highly sensitive, genetically manipulated mouse models has great potential to improve understanding and quantification of risk at low doses, such as the dose received in occupational and medical exposures.
REFERENCES 1. Bouffler S, Silver A, Cox R. Mechanistic and genetic studies of radiation tumorigenesis in the mouse—implications for low dose risk estimation. J Radiol Prot 2002;22:A11–A16. 2. Di Majo V, Rebessi S, Pazzaglia S, Saran A, Covelli V. Carcinogenesis in laboratory mice after low doses of ionizing radiation. Radiat Res 2003;159:102–108. 3. Goodrich LV, Milenkovic´ L, Higgins KM, Scott MP. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997;277:1109–1113. 4. Hahn H, Wojnowski L, Zimmer AM, et al. Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlin syndrome. Nat Med 1998;4:619–622. 5. Aszterbaum M, Epstein J, Oro A, et al. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice. Nat Med 1999;5: 1285–1291. 6. Mancuso M, Pazzaglia S, Tanori M, et al. Basal cell carcinoma and its development: Insights from radiation-induced tumors in Ptch1-deficient mice. Cancer Res 2004;64:934–941. 7. Hahn H, Nitzki F, Schorban T, et al. Genetic mapping of a Ptch1-associated rhabdomyosarcoma susceptibility locus on mouse chromosome 2. Genomics 2004;84:853–858. 8. Pazzaglia S, Tanori M, Mancuso M, et al. Linking DNA damage to medulloblastoma tumorigenesis in patched heterozygous knockout mice. Oncogene 2006;25:1165–1173. 9. Pazzaglia S, Mancuso M, Atkinson MJ, et al. High incidence of medulloblastoma following X-ray-irradiation of newborn Ptc1 heterozygous mice. Oncogene 2002;21:7580–7584. 10. Mancuso M, Pasquali E, Leonardi S, et al. Oncogenic bystander radiation effects in Patched heterozygous mouse cerebellum. Proc Natl Acad Sci U S A 2008;105:12445–12450.
11. Calzada-Wack J, Kappler R, Schnitzbauer U, et al. Unbalanced overexpression of the mutant allele in murine Patched mutants. Carcinogenesis 2002;23:727–733. 12. Wechsler-Reya R, Scott MP. The developmental biology of brain tumors. Annu Rev Neurosci 2001;24:385–428. 13. Pazzaglia S, Tanori M, Mancuso M, et al. Two-hit model for progression of medulloblastoma preneoplasia in Patched heterozygous mice. Oncogene 2006;25:5575–5580. 14. Pazzaglia S, Mancuso M, Tanori M, et al. Modulation of patchedassociated susceptibility to radiation induced tumorigenesis by genetic background. Cancer Res 2004;64:3798–3806. 15. Mitchel REJ. Radiation risk prediction and genetics: The influence of the TP53 gene in vivo. Dose-Response 2005;3:519–532. 16. Wakeford R. The cancer epidemiology of radiation. Oncogene 2004;23:6404–6428. 17. Preston DL, Ron E, Yonehara S, et al. Tumors of the nervous system and pituitary gland associated with atomic bomb radiation exposure. J Natl Cancer Inst 2002;94:1555–1563. 18. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) 2000. Sources and effects of ionizing radiation. Report to the General Assembly with Annexes. Vol. II. New York: United Nations. 19. Ron E, Modan B, Boice JD Jr., et al. Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 1988;319:1033–1039. 20. Karlsson P, Holmberg E, Lundell M, et al. Intracranial tumors after exposure to ionizing radiation during infancy: A pooled analysis of two Swedish cohorts of 28,008 infants with skin hemangioma. Radiat Res 1998;150:357–364. 21. Little MP, de Vathaire F, Shamsaldin A, et al. Risks of brain tumour following treatment for cancer in childhood: Modification
1210
I. J. Radiation Oncology d Biology d Physics
by genetic factors, radiotherapy and chemotherapy. Int J Cancer 1998;78:269–275. 22. Hawkins MM, Draper GJ, Kingston JE. Incidence of second primary tumours among childhood cancer survivors. Br J Cancer 1987;56:339–347. 23. Walter AW, Hancock ML, Pui CH, et al. Secondary brain tumors in children treated for acute lymphoblastic leukemia at St Jude Children’s Research Hospital. J Clin Oncol 1998;16: 3761–3767.
Volume 73, Number 4, 2009
24. Neglia JP, Robison LL, Stovall M, et al. New primary neoplasms of the central nervous system in survivors of childhood cancer: A report from the Childhood Cancer Survivor Study. J Natl Cancer Inst 2006;98:1528–1537. 25. Travis LB. The epidemiology of second primary cancers. Cancer Epidemiol Biomarkers Prev 2006;15:2020–2026. 26. Flint-Richter P, Sadetzki S. Genetic predisposition for the development of radiation-associated meningioma: An epidemiological study. Lancet Oncol 2007;8:403–410.
doi:10.1016/j.ijrobp.2008.10.068
BIOLOGY CONTRIBUTION
PHYSICAL, HERITABLE AND AGE-RELATED FACTORS AS MODIFIERS OF RADIATION CANCER RISK IN PATCHED HETEROZYGOUS MICE SIMONETTA PAZZAGLIA, PH.D.,* EMANUELA PASQUALI, M.SC.,y MIRELLA TANORI, PH.D.,* MARIATERESA MANCUSO, PH.D.,* SIMONA LEONARDI, PH.D.,* VINCENZO DI MAJO, PH.D.,* SIMONETTA REBESSI, PH.D.,* AND ANNA SARAN, PH.D.* * Section of Toxicology and Biomedical Sciences, Biotechnologies, Agro-Industry and Health Protection Department, ENEA CR Casaccia, 00123 Rome, Italy, and y Department of Experimental Oncology, Istituto Nazionale Tumori, 20133 Milan, Italy Purpose: To address the tumorigenic potential of exposure to low/intermediate doses of ionizing radiation and to identify biological factors influencing tumor response in a mouse model highly susceptible to radiogenic cancer. Methods and Materials: Newborn Ptc1 heterozygous mice were exposed to X-ray doses of 100, 250, and 500 mGy, and tumor development was monitored for their lifetime. Additional groups were irradiated with the same doses and sacrificed at fixed times for determination of short-term endpoints, such as apoptosis and early preneoplastic lesions in cerebellum. Finally, groups of Ptc1 heterozygous mice were bred on the C57BL/6 background to study the influence of common variant genes on radiation response. Results: We have identified a significant effect of low-intermediate doses of radiation (250 and 500 mGy) in shortening mean survival and inducing early and more progressed stages of tumor development in the cerebellum of Ptc1+/– mice. In addition, we show that age at exposure and heritable factors are potent modifiers of radiationrelated cancer risk. Conclusions: The Ptc1 knockout mouse model offers a highly sensitive system that may potentially help to improve understanding and quantification of risk at low doses, such as doses experienced in occupational and medical exposures, and clarify the complex interactions between genetic and environmental factors underlying cancer susceptibility. Ó 2009 Elsevier Inc. Patched1, Medulloblastoma, Cancer risk, Genetic background, Ionizing radiation.
Animal studies play an important role in improving our understanding of radiation carcinogenesis (1, 2). The use of tumor data from animal studies is necessary as a complement of epidemiological studies of human populations to develop estimates of radiation cancer risk at low doses. In addition, animal experiments provide valuable insights into the mechanisms of radiation interaction with living cells and organisms, allowing clarification of the pathways of tumorigenesis and of the factors modifying radiation risks. Animal studies, however, are hampered by the requirement of very large animal numbers to reach statistical significance, particularly at low radiation doses. Thus, genetically manipulated, radiation-susceptible mouse models represent a powerful tool to help assessment of risk. In addition, a strong need exists for systems that provide information on the mechanisms whereby a ‘‘hit’’ normal cell develops into a tumor.
The Ptc1 knockout mouse model is a well-established model of carcinogenesis, particularly in the brain, skin, and soft tissues (3–5). Radiation is known to elevate cerebellum tumor incidence and to promote basal cell carcinoma (BCC) precursor lesions to progress to large infiltrative BCC (6), but little is known of low-dose radiation effects or the shape of the dose–response for radiation-induced cancers at low doses. Understanding of other factors that are known to modify radiation-related tumorigenesis is also incomplete. To investigate cancer susceptibility to low to intermediate doses, we carried out experiments for assessment of tumor responses to 100, 250, and 500 mGy of X-rays in Ptc1+/– mice. Dose dependence was also established for several biological endpoints such as tumor-free survival, apoptotic cell death, and early-preneoplastic areas in cerebellum. Age at irradiation and genetic background were also examined as modifiers of radiation-related cancer risk in cerebellum.
Reprint requests to: Simonetta Pazzaglia, Ph.D., Section of Toxicology and Biomedical Sciences, Biotechnologies, Agro-Industry and Health Protection Department, ENEA CR Casaccia, 00123 Rome, Italy. Tel: (+39) 06-3048-6535; Fax: (+39) 06-3048-3644; E-mail: [email protected]
Conflict of interest: none Acknowledgments—This work was supported by EU Contract FI6RCT-2003-508842 RISC-RAD. Received Aug 5, 2008, and in revised form Oct 15, 2008. Accepted for publication Oct 16, 2008.
INTRODUCTION
1203
1204
I. J. Radiation Oncology d Biology d Physics
Analysis of Ptc1 LOH
METHODS AND MATERIALS Mice neo67/+
Volume 73, Number 4, 2009
+/–
Mice lacking one Ptc1 allele (Ptc , termed Ptc1 throughout the text), generated through disruption of exons 6 and 7 in 129S2 ES cells (4) and maintained on a CD1 or C57BL/6 background, were used in this study. CD1 is an outbred stock, and brother-sister mating was avoided to minimize consanguinity. C57BL/6-Ptc1+/– mice were generated by breeding the 129S2 mutation for at least 10 generations onto the C57BL/6NCrlBR (B6) mice purchased from Charles River Laboratorie (Sulzfeld, Germany) (7). Mice were genotyped using polymerase chain reaction primers specific to the neo insert and wildtype regions as described previously (4). Animals were housed under conventional conditions with food and water available ad libitum and a 12-h light-dark cycle. Experimental protocols were reviewed by the Internal Institutional Animal Care and Use Committee.
Irradiation For tumor induction, groups of Ptc1+/– mice on CD1 background were whole-body irradiated at postnatal Day 1 (P1) with 100 (n = 45), 250 (n = 45), or 500 mGy (n = 39), or left untreated (n = 51). For study of genetic background effects, Ptc1+/– mice on C57BL/ 6 background were whole-body irradiated at P1 with 3 Gy of X-rays (n = 15) or left unirradiated (n = 22). For age-at-irradiation effects, data from previous experiments with X-rays (3 Gy) delivered at P1, P2, P4, and P10 were reanalyzed (8–10). Irradiation was performed using a Gilardoni CHF 320 G X-ray generator (Gilardoni, Mandello del Lario, Lecco, Italy) operated at 250 kVp, 1 mA for 100 mGy (dose rate: 59 mGy/min), 5 mA for 250 and 500 mGy (dose rate: 327 mGy/min), 15 mA for 3Gy (dose rate: 1,000 mGy/ min), with filters of 2.0 mm Al and 0.5 mm Cu (HVL=1.6 mm Cu).
Tumor quantification and histological analysis Mice were observed daily for their whole life span. Upon decline of health (i.e., severe weight loss, paralysis, ruffling of fur, or inactivity), they were euthanized and necropsied. Whole brain, dorsal skin, and any visible mass were fixed in 4% buffered formalin. Samples were then embedded in paraffin wax according to standard techniques. Cerebellar tumors and normal tissue were snap frozen. Tumor incidence was expressed as the percentage of mice with one or more tumors. Incidence of microscopic cerebellar lesions was determined on asymptomatic Ptc1+/– mice irradiated with 100 (n = 17), 250 (n = 12), or 500 mGy (n = 14) X-rays at P1, or left untreated (n = 16), and sacrificed at 8 weeks postirradiation. In total, 18 sections were examined for each cerebellum with an interval of 70 mm.
Radiosensitivity analysis Brains (n = 3 per time point) of irradiated and unirradiated Ptc1+/– pups were collected at 6 h postirradiation and fixed in 4% buffered formalin. Serial sections of cerebellar tissues were cut at 4-mm thickness and stained with hematoxylin and eosin. Digital images from midsagittal cerebellar sections of P1 pups, covering the entire cerebellar surface, were collected by IAS image-processing software (Delta Sistemi, Rome, Italy). Cells showing signs of nuclear chromatin condensation and morphologically normal cells in the external granule layer (EGL) of the cerebellum were counted using a double-blind method. Apoptotic values were calculated as the percentage of pyknotic nuclei relative to the total number of cells. The total number of cells examined ranged from 0.9 104 to 1.4 104.
Genomic and tumor DNA was extracted from spleens and medulloblastomas by Wizard SV Genomic DNA Purification System (Promega, Madison, WI). LOH analysis at the Ptc1 locus was performed by amplification and sequencing of exon 23 of the murine Ptc1 gene, which holds a polymorphic site at position 4016 between the targeted Ptc1 allele in the 129S2 strain and the wildtype allele from the CD1 stock. The primers used for the study were primer pair 8F/9R (11). Sequencing reactions were performed by means of dye terminator chemistry using a Big Dye Terminator version 3.1 Sequencing Cycle Kit and a 3730 DNA Analyzer (Applied Biosystems, Foster City, CA). Loss of polymorphic sites in medulloblastoma compared with normal tissue was considered LOH at the Ptc1 locus.
Statistics Survival results are reported as means SE; Mann-Whitney U test was used for determination of statistical differences between groups. Kaplan-Meier medulloblastoma-free survival curves were compared, and log-rank test p values were calculated. Analyses were done using GraphPad Prism version 4.02 for Windows (GraphPad Software, San Diego, CA). A p value of < 0.05 was considered statistically significant.
RESULTS Survival and pathology of Ptc1+/– mice after low-dose X-ray irradiation The tumor-free survival of Ptc1+/– mice exposed to 0, 100, 250, and 500 mGy of X-rays at P1 is shown in Fig. 1a. For comparison, survival of mice irradiated with 3 Gy is reported from previous work (7). Survival data are summarized in Fig. 1b. The mean survival was significantly reduced in Ptc1+/– mice exposed to 250 mGy (46.4 4.6; p = 0.0451) and 500 mGy (37.4 3.81; p = 0.0007) compared with unirradiated mice (56.0 3.7). Reduction in survival was also observed in Ptc1+/– mice irradiated with the lowest 100 mGy dose (48.8 3.3), although this was not statistically significant. The tumor spectrum of the different Ptc1+/– groups is shown in Fig. 2b. Unexposed mice showed the expected rates of spontaneously occurring Ptc1-associated tumors, i.e., medulloblastomas and soft tissue sarcomas. In irradiated groups, medulloblastomas occurred at an enhanced rate with increasing dose level and represented a competing cause of death relative to other tumors. There was a fourfold increase in medulloblastoma incidence in mice irradiated with 250 mGy (p = 0.0113), and a sixfold increase at 500 mGy (p < 0.0001) relative to controls. In mice exposed to 100 mGy, we found a nearly twofold increase over controls; however, this difference was not statistically significant, and a larger group size (i.e., n > 200) would be required. Although noninducible sarcomas remained the predominant tumor type in Ptc1+/– mice exposed to 100 and 250 mGy (47.6% and 35.7%), medulloblastoma was the most frequent tumor (50%) at 500 mGy. Tumors started between 8 and 12 weeks of age in all groups (Fig. 2a). Compared with previous experiments at 3 Gy (Fig. 2a, dashed line; median survival = 11 weeks) (8),
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
(a) 100 100 mGy 250 mGy
80
500 mGy 3000 mGy
60 40 20
Medulloblastoma free mice (%)
(a) 0 mGy
Percent survival
1205
100
0 mGy 100 mGy
80
250 mGy 60
500 mGy 3000 mGy
40 20 0 0
10
20
30
40 80 100
Time (weeks)
0 0
25
50
75
100
(b)
Weeks of age
(b) Dose (mGy)
ap
Number of mice
Mean survival (weeks ± SE)
0
51
100
45
56.2 ± 3.7 48.8 ± 3.3
250
45
46.4 ± 4.6a
500
39
37.4 ± 3.8b
= 0.0451; b p = 0.0007 in the Mann-Whitney U test
Fig. 1. Tumor-free survival of Ptc1+/– mice. (a) Survival of Ptc1+/– mice exposed to 0, 100, 250, and 500 mGy of X-rays at P1. Survival of mice irradiated with 3 Gy is shown for comparison (dashed line) (8). (b) Mean survival after exposure to X-rays (100, 250, or 500 mGy) compared with control Ptc1+/– mice.
the time course of medulloblastoma development was significantly delayed at 500 mGy (median survival = 39 weeks; p < 0.0001, log-rank test). Whereas 50% of the tumors (18/36) from the current experiments developed in mice older than 20 weeks, all medulloblastomas (17/17) from the 3-Gy experiment arose within 20 weeks postirradiation, and 60% of the animals had developed tumors by week 13. These results indicate that the kinetics of medulloblastoma development can be significantly influenced depending on radiation dose. The skin represents an additional target of ionizing radiation in the Ptc1+/– model because exposure to 3 Gy results in significant induction of BCC (6). Notably, Ptc1+/– mice irradiated with 100–500 mGy and observed for their lifetime did not develop skin tumors, revealing substantial differences in radiosensitivity between skin and cerebellum.
Quantitative dose–response relationships In Ptc1 heterozygous mice, we examined the dose– response curves in the low-dose range for several biological endpoints, including tumor-free survival, apoptotic cell death in the EGL of the cerebellum, early preneoplastic cerebellar lesions, and full medulloblastoma development. Tumor-free survival. The survival of irradiated Ptc1+/– mice in the dose range of 0–500 mGy is shown in Fig. 3. Data are well fitted using a nonthreshold linear dose dependence (r2 = 0.87). When the larger 0–3,000 mGy dose range was considered, a linear-quadratic dose dependence represented the best fit (r2 = 0.98; Fig. 3, inset).
Tumor type (%)
0 mGy
100 mGy
250 mGy 500 mGy
Medulloblastoma BCC Lymphoma Sarcoma Others
4 (7.8) 2 (3.9) 19 (37.3) 6 (11.8)
6 (14.3) 3 (7.1) 20 (47.6) 5 (11.9)
12 (28.6)a 18 (50.0)b 1 (2.4) 1 (2.8) 15 (35.7) 8 (22.2) 4 (9.52) 2 (5.5)
No of mice with tumors
31 (60.8)
31 (73.8)
29 (69.0) 28 (77.8)
51
42
No of autopsied mice
42
36
a p = 0.0113; b p<0.0001 in the log rank test
Fig. 2. Rate of medulloblastoma development after low-dose irradiation in Ptc1+/– mice. (a) Kaplan-Meier kinetic analysis of medulloblastoma in Ptc1+/– mice exposed to 100, 250, and 500 mGy of X-rays. Accelerated medulloblastoma rate in mice irradiated with 3 Gy is shown for comparison (dashed line; ref. 8). (b) Tumor spectrum in irradiated (100, 250, or 500 mGy) and control Ptc1+/– mice.
Radiation-induced cell death in granule cell progenitors (GCPs). GCPs are considered the cells of origin of medulloblastoma (12). At P1, proliferating GCPs are clustered over the surface of the developing cerebellum to form the EGL. By quantification of pyknotic nuclei in the EGL of P1-cerebella, we examined the apoptotic levels at 6 h after exposure to 100, 250, and 500 mGy. Significant differences
Fig. 3. Dose–effect relationship for mean survival time in Ptc1+/– mice exposed to low X-ray doses at P1. The fit was linear in the dose range 0–500 mGy. (Inset) Data were best fitted by a linear-quadratic dose dependence in the 0–3,000 mGy dose interval. Data points are means standard error.
1206
I. J. Radiation Oncology d Biology d Physics
Volume 73, Number 4, 2009
were statistically significant between Ptc1+/– mice irradiated with 250 (p = 0.0242) or 500 mGy (p = 0.0051) and unirradiated mice. The increase was not significant in mice irradiated with 100 mGy. We fit these data using a linear regression (r2 = 0.70; slope = 0.1). To test whether other models would provide a better fit, we tested a linear-quadratic model. This improved fit (r2 = 0.9). However, when only the 100 and 250 mGy data points were considered, the data showed a significant linear trend (r2 = 0.88). Data in the 0–3,000 mGy dose interval were best fitted using a linear-quadratic dose dependence (r2 = 0.90; Fig. 5C, inset). Analysis of medulloblastoma development in cerebellum of Ptc1+/– mice showed clear dose dependence at doses below 500 mGy (Fig. 5D and 5E). A very good linearity (r2 = 0.99) and a similar slope (i.e., 0.086) compared with the early-lesion endpoint (slope ratio = 1.2) suggest similar effectiveness of radiation for induction of preneoplastic and fully malignant growth. Data relative to the 0–3,000 mGy range were best fitted using a linear-quadratic dose dependence (r2 = 0.99; Fig. 5F, inset).
+/–
Fig. 4. Apoptotic response in the cerebellum of Ptc1 mice exposed to low X-ray doses at P1. (A) Hematoxylin and eosin stained sections of cerebellum at 6 h after irradiation with 100 mGy, (B) 250 mGy, (C) 500 mGy, and (D) unirradiated cerebellum. (E) Dose response for apoptosis in the external granule layer (EGL) of Ptc1+/– mice after irradiation with 100, 250, and 500 mGy of X-rays. Data in the interval 0–500 mGy were fitted with a linear regression. The best fit of the data in the 0–3,000 mGy dose interval is represented by a linear-quadratic dose dependence (inset). Data points represent the percentage of apoptotic cells out of the total number of cells in the EGL standard error.
in the number of cells undergoing apoptosis in irradiated compared with control cerebellum were detected at all doses (Fig. 4 A–D; p < 0.0001). Incidences were 2.1, 4.1, and 10.7% of GCPs at 100, 250, and 500 mGy, respectively. A linear regression yielded a good fit for the kinetics of cell inactivation over this range (r2 = 0.98; Fig. 4E). The best data fit was again represented by a linear-quadratic dose dependence (r2 = 0.99) when data in the 0–3,000 mGy range were analyzed (Fig. 4E, inset). Dose responses for early preneoplastic EGL lesions and medulloblastoma induction. Young Ptc1+/– mice develop a preneoplastic phenotype represented by ectopic EGL regions in the cerebellum of asymptomatic mice (8). We have shown previously that exposure to 3 Gy of X-rays substantially increases the size and frequency of such lesions (13), but the effects of low to intermediate doses have not been assessed. We performed histological examination of cerebella at 8 weeks after neonatal irradiation with 100, 250, and 500 mGy (Fig. 5A and 5B) and show that the frequency of cerebellar abnormalities in asymptomatic mice increases with radiation dose in the 0–500 mGy dose range (Fig. 5C). Differences in the frequency of cerebellar lesions
Ptc1 loss of heterozygosity (LOH) in medulloblastomas induced by low-dose exposure Biallelic Ptc1 loss represents an early and critical event in medulloblastoma development in Ptc1+/– mice exposed to radiation and the biological switch to malignancy in early cerebellar lesions (13). We asked whether the processes that drive carcinogenesis due to high dose exposures also contribute to low-dose radiation carcinogenesis. To investigate whether radiation-induced DNA damage and tumorigenic mechanisms differ at low doses, we assayed loss of the wildtype Ptc1 allele in medulloblastomas induced at doses below 500 mGy. Sequence analysis of DNA showed that, similar to tumors induced at higher dose, lack of the wildtype Ptc1 allele was found in all (9/9) medulloblastomas analyzed (Fig. 5 E). Age-at-irradiation effects We have previously shown that ionizing radiation induces medulloblastoma only in neonatal Ptc1+/– mice (9). The age– response relationship for tumor induction in the cerebellum after exposure to 3 Gy of X-rays at P1, P2, P4, and P10 is shown in Fig. 6a. There was a drastic decrease in tumor incidence with increasing age at irradiation from P1 to P10, age at which tumorigenic effects of radiation decreased out despite dramatically increasing size of cerebellum, and number of target cells. Sagittal sections of cerebella at P1, P2, P4, and P10 are shown in Fig. 6 b–e. The remarkable decrease in sensitivity with increasing age at irradiation suggests that developmental changes occurring after birth diminish sensitivity to radiation-induced malignancy in the central nervous system (CNS). Genetic factors In Ptc1 mice, genetic background strongly affects the spectrum of neoplasms (4, 7, 14). To test whether modifiers of
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
1207
Fig. 5. Tumorigenesis by low radiation doses in the cerebellum of Ptc1+/– mice. (A and B) Histological features of early cerebellar lesions (arrows) in Ptc1+/– mice irradiated at P1. (C) Dose–effect relationship for the induction of early cerebellar lesions in Ptc1+/– mice irradiated with 100, 250, and 500 mGy of X-rays; data were fitted by a linear dose-dependence (r2 = 0.7). (Inset) Best fit of the data in the 0–3,000 mGy dose interval. (D) Gross appearance of a normal brain from Ptc1 mice and relative DNA sequencing showing the presence of a polymorphism (T/C) at position 4016 of the Ptc1 gene; this site differs between the 129S2 strain (T allele), on which the Ptc1 knock out was engineered, and the CD-1 stock which is the source of the wildtype Ptc1 (C allele). (E) Brain with a large medulloblastoma (arrow) and electropherogram of tumor DNA showing only the targeted allele and demonstrating loss of the wildtype Ptc1 allele. (F) Dose response for full medulloblastoma development after irradiation with 100, 250, and 500 mGy of X-rays. (Inset) A linear quadratic dose dependence represents the best fit in the 0–3,000 mGy dose interval.
cancer susceptibility also modify radiation-induced cancer, neonatal Ptc1+/– mice on C57BL/6 background were exposed to 3 Gy of X-rays. We observed a spontaneous sevenfold higher incidence of medulloblastoma (40%) in C57BL/ 6-Ptc1+/– compared with CD1-Ptc1+/– mice (7.7%; p = 0.001, chi-square test; Fig. 7a and 7b). Notably, no induction of medulloblastoma over the spontaneous rate was observed in C57BL/6-Ptc1+/– mice after irradiation with 3 Gy (Fig. 7b). For comparison, the tumor response in neonatally irradiated CD1-Ptc1+/– mice is reported from a previous publication (8) (Fig. 7a). These findings suggest that gene–environment interactions are important determinants in cancer risks from radiation.
DISCUSSION Cancer risk caused by exposure to low doses of ionizing radiations is particularly important for assessing the safety of clinical and industrial applications of radiation. In this study, a mouse model of spontaneous and radiogenic cancer has been used to investigate the effects of exposure to doses below 500 mGy on tumor yields. We have been able to detect significant increases in brain tumor incidence at doses of 250 and 500 mGy of X-rays. In contrast, the skin, an additional radiation target in the Ptc1+/– model, was not susceptible to radiation doses below 500 mGy. Thus, although Ptc1-associated BCC and medulloblastoma show common features (i.e.,
I. J. Radiation Oncology d Biology d Physics
1208
Volume 73, Number 4, 2009
The experiments described here were designed to test whether the dose–response relationships for several parameters (i.e., life shortening, induction of apoptosis, and cancer induction) could be determined by limited-size animal experiments and to assess the shape of the dose–response curve. Experimental data from this study provide clear quantitative evidence that a linear dose dependence is adequate to describe responses for several biological endpoints in the dose range 0–500 mGy. By contrast, a linear-quadratic dependence represented the best fit for all endpoints when experimental data in the dose range 0–3,000 mGy were considered. Similar to previous experiments at 3 Gy, the same molecular event (i.e., specific loss of the wildtype Ptc1 allele) was detected in medulloblastomas induced by low radiation doses, showing that radiation is acting by inducing gene loss events. However, medulloblastoma development was significantly delayed at doses # 500 mGy, suggesting that radiation dose influences the time frame for Ptc1 LOH. For radiological protection, it is important to gain a better view of the common factors influencing radiation response, and the age at irradiation is known to influence cancer risk (16). Data on the effect of postnatal age at irradiation from follow-up studies of atomic bomb survivors show that relative cancer risks are higher at younger ages for several cancer types (17, 18). Our results provide quantitative evidence that neonatal Ptc1+/– mice are at greatest risk of radiation-associated brain cancer, possibly reflecting a larger proportion of dividing cells that could acquire tumor-initiating mutations in early postnatal cerebellum (19). Strong similarities exist between brain developmental processes in mice and humans. In mice, the formation of cerebellum spans embryonic and postnatal development, initiating at 9 days postcoitus and completing by postnatal Day 20. In humans the nervous system begins to form during the first trimester of gestation and protracts until after birth with extensive postnatal neurogenesis occurring in the cerebellum for a further 1 year. Compared with mouse, one might expect increased radiation
Fig. 6. Age-at-exposure effects for medulloblastoma (Mb) tumorigenesis after irradiation of Ptc1+/– mice with 3 Gy of X-rays. (a) Decrease in tumor incidence with increasing age at irradiation from P1 to P10. (b–e) Hematoxylin and eosin stained sections of mouse cerebellum at P1 (b), P2 (c), P4 (d), and P10 (e), showing progressive increase in the size during early postnatal development.
both are radiogenic tumors developing through early preneoplastic lesions that require biallelic Ptc1 loss for progression to full malignancy) (6, 13), skin and cerebellum show markedly different radiosensitivity to tumor induction. The basic concept that individual tissues differ in quantitative response to radiation also emerges in a study on cancer-prone TP53 heterozygous mice in which a single exposure to 100 mGy of X-rays decreased risk of lymphoma by extending tumor latency but increased risk of osteosarcoma (15). Because most environmental exposures occur at low doses, one of the main issues in radioprotection is to assess the shape of the dose–response curve in the low-dose region.
(a)
(b)
Medulloblastoma incidence( )
3 Gy (n=21)
100
CD1 0 Gy (n=51)
3 Gy (n=15)
100
C57BL/6 0 Gy (n=22)
75
75
50
50
25
25 0
0 0
10
20
30
Age (weeks)
40
50
0
10
20
30
40
50
Age (weeks)
Fig. 7. Influence of genetic background on development of spontaneous and radiogenic medulloblastoma in Ptc1+/– mice. (a) Neonatal irradiation of CD1-Ptc1+/– mice with 3 Gy of X-rays induced a 10-fold increase in medulloblastoma final rate relative to controls (8). (b) Irradiation of neonatal C57BL/6-Ptc1+/– mice with the same dose did not enhance medulloblastoma. A sevenfold higher incidence of spontaneous medulloblastoma was observed in C57BL/6-Ptc1+/– mice compared with CD1-Ptc1+/– mice (p = 0.001, chi-square test).
Low-dose radiobiology in Ptc1+/– mice d S. PAZZAGLIA et al.
risk of medulloblastoma for children younger than 1 year. Importantly, exposure to radiation therapy has been recognized as the most important risk factor for the development of a new CNS tumor in survivors of childhood cancers (19–23). In particular, excess relative risk is highest among children exposed at younger than 5 years, suggesting a greater susceptibility of the developing brain to radiation (24). Our tumor data in mice confirm that the age at exposure is a consistent modifier of radiation-related CNS cancer risk and that significant differences exist between the developing and mature nervous system in tumor response to exogenous DNA damage. Currently, within age and sex groups, radiation risk is assumed to be uniformly distributed across populations. However, genetic variations may cause individual variability in biological responses to radiation and increased risk in susceptible individuals (25). In the presence of high penetrance alleles, genetics plays a major role in determining individual risk. The Ptc1 gene, indeed, represents a paradigm of radiation hypersensitivity, leading to cancer. However, frequencies of such high penetrance alleles in humans are low, and greater contribution to radiation risks is probably due to genes that are less penetrant but more frequent. We show here that the genetic background hosting the Ptc1 mutation can dramatically alter the individual risk for developing a specific tumor. In fact, whereas spontaneous medulloblastoma was significantly increased when the Ptc1+/– mutation was expressed in the C57BL/6 background, C57BL/6-Ptc+/–
1209
mice were completely resistant to medulloblastoma induction by radiation, suggesting that different sets of genes may control susceptibility to spontaneous and radiation-induced cancer. These findings indicate that complex interactions between genetic and environmental factors underlie susceptibility to radiogenic cancer in the CNS. Interestingly, familial aggregation for development of radiation-associated meningioma has been recently shown, supporting the hypothesis of genetic predisposition for radiogenic CNS tumors in humans (26). Thus, a better view of the common variant genes that influence radiation response would contribute to the development of more refined approaches to assessment of radiation health risk in humans. In summary, we have employed a mouse model of radiogenic cancer to assess the tumorigenic potential of exposure to low to intermediate doses of ionizing radiation and to identify biological factors influencing tumor response. Our in vivo findings identify a linear dose–response relationship in the low-dose region for several biological endpoints, such as tumor-free survival, apoptosis, and tumor induction. We also provide evidence that age at exposure and heritable factors are potent modifiers of radiation-related cancer risk in a susceptible mouse model. Although a genetically susceptible population represents a worst-case scenario, the use of highly sensitive, genetically manipulated mouse models has great potential to improve understanding and quantification of risk at low doses, such as the dose received in occupational and medical exposures.
REFERENCES 1. Bouffler S, Silver A, Cox R. Mechanistic and genetic studies of radiation tumorigenesis in the mouse—implications for low dose risk estimation. J Radiol Prot 2002;22:A11–A16. 2. Di Majo V, Rebessi S, Pazzaglia S, Saran A, Covelli V. Carcinogenesis in laboratory mice after low doses of ionizing radiation. Radiat Res 2003;159:102–108. 3. Goodrich LV, Milenkovic´ L, Higgins KM, Scott MP. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997;277:1109–1113. 4. Hahn H, Wojnowski L, Zimmer AM, et al. Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlin syndrome. Nat Med 1998;4:619–622. 5. Aszterbaum M, Epstein J, Oro A, et al. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice. Nat Med 1999;5: 1285–1291. 6. Mancuso M, Pazzaglia S, Tanori M, et al. Basal cell carcinoma and its development: Insights from radiation-induced tumors in Ptch1-deficient mice. Cancer Res 2004;64:934–941. 7. Hahn H, Nitzki F, Schorban T, et al. Genetic mapping of a Ptch1-associated rhabdomyosarcoma susceptibility locus on mouse chromosome 2. Genomics 2004;84:853–858. 8. Pazzaglia S, Tanori M, Mancuso M, et al. Linking DNA damage to medulloblastoma tumorigenesis in patched heterozygous knockout mice. Oncogene 2006;25:1165–1173. 9. Pazzaglia S, Mancuso M, Atkinson MJ, et al. High incidence of medulloblastoma following X-ray-irradiation of newborn Ptc1 heterozygous mice. Oncogene 2002;21:7580–7584. 10. Mancuso M, Pasquali E, Leonardi S, et al. Oncogenic bystander radiation effects in Patched heterozygous mouse cerebellum. Proc Natl Acad Sci U S A 2008;105:12445–12450.
11. Calzada-Wack J, Kappler R, Schnitzbauer U, et al. Unbalanced overexpression of the mutant allele in murine Patched mutants. Carcinogenesis 2002;23:727–733. 12. Wechsler-Reya R, Scott MP. The developmental biology of brain tumors. Annu Rev Neurosci 2001;24:385–428. 13. Pazzaglia S, Tanori M, Mancuso M, et al. Two-hit model for progression of medulloblastoma preneoplasia in Patched heterozygous mice. Oncogene 2006;25:5575–5580. 14. Pazzaglia S, Mancuso M, Tanori M, et al. Modulation of patchedassociated susceptibility to radiation induced tumorigenesis by genetic background. Cancer Res 2004;64:3798–3806. 15. Mitchel REJ. Radiation risk prediction and genetics: The influence of the TP53 gene in vivo. Dose-Response 2005;3:519–532. 16. Wakeford R. The cancer epidemiology of radiation. Oncogene 2004;23:6404–6428. 17. Preston DL, Ron E, Yonehara S, et al. Tumors of the nervous system and pituitary gland associated with atomic bomb radiation exposure. J Natl Cancer Inst 2002;94:1555–1563. 18. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) 2000. Sources and effects of ionizing radiation. Report to the General Assembly with Annexes. Vol. II. New York: United Nations. 19. Ron E, Modan B, Boice JD Jr., et al. Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 1988;319:1033–1039. 20. Karlsson P, Holmberg E, Lundell M, et al. Intracranial tumors after exposure to ionizing radiation during infancy: A pooled analysis of two Swedish cohorts of 28,008 infants with skin hemangioma. Radiat Res 1998;150:357–364. 21. Little MP, de Vathaire F, Shamsaldin A, et al. Risks of brain tumour following treatment for cancer in childhood: Modification
1210
I. J. Radiation Oncology d Biology d Physics
by genetic factors, radiotherapy and chemotherapy. Int J Cancer 1998;78:269–275. 22. Hawkins MM, Draper GJ, Kingston JE. Incidence of second primary tumours among childhood cancer survivors. Br J Cancer 1987;56:339–347. 23. Walter AW, Hancock ML, Pui CH, et al. Secondary brain tumors in children treated for acute lymphoblastic leukemia at St Jude Children’s Research Hospital. J Clin Oncol 1998;16: 3761–3767.
Volume 73, Number 4, 2009
24. Neglia JP, Robison LL, Stovall M, et al. New primary neoplasms of the central nervous system in survivors of childhood cancer: A report from the Childhood Cancer Survivor Study. J Natl Cancer Inst 2006;98:1528–1537. 25. Travis LB. The epidemiology of second primary cancers. Cancer Epidemiol Biomarkers Prev 2006;15:2020–2026. 26. Flint-Richter P, Sadetzki S. Genetic predisposition for the development of radiation-associated meningioma: An epidemiological study. Lancet Oncol 2007;8:403–410.