Sparing and enhancing dose protraction effects for radiation damage to the aorta of wild-type mice

Abstract Purpose Our previous work indicated the greater magnitude of damage to the thoracic aorta at 6 months after starting 5 Gy irradiation in descending order of exposure to X-rays in 25 fractions > acute X-rays > acute γ-rays > X-rays in 100 fractions ≫ chronic γ-rays, in which the limitations of the study included a lack of data for fractionated γ-ray exposure. To better understand effects of dose protraction and radiation quality, the present study examined changes after exposure to γ-rays in 25 fractions, and compared its biological effectiveness with five other irradiation regimens. Materials and Methods Male C57BL/6J mice received 5 Gy of 137Cs γ-rays delivered in 25 fractions spread over six weeks. At 6 months after starting irradiation, mice were subjected to echocardiography, followed by tissue sampling. The descending thoracic aorta underwent scanning electron microscopy, immunofluorescence staining and histochemical staining. The integrative analysis of multiple aortic endpoints was conducted for inter-regimen comparisons. Results Exposure to γ-rays in 25 fractions induced vascular damage (evidenced by increases in endothelial detachment and vascular endothelial cell death, decreases in endothelial waviness, CD31, endothelial nitric oxide synthase and vascular endothelial cadherin), inflammation (evidenced by increases in tumor necrosis factor α, CD68 and F4/80) and fibrosis (evidenced by increases in transforming growth factor β1, alanine blue stain and intima-media thickness). The integrative analysis revealed biological effectiveness in descending order of exposure to X-rays in 25 fractions > acute X-rays > γ-rays in 25 fractions > acute γ-rays > X-rays in 100 fractions ≫ chronic γ-rays. Conclusions The results suggest that dose protraction effects on aortic damage depend on radiation quality, and are not a simple function of dose rate and the number of fractions.


Introduction
In occupational settings and during medical procedures, ionizing radiation exposure of the circulatory system occurs over a wide range of doses and dose rates.A growing body of epidemiological evidence for increased radiation risks of diseases of the circulatory system (DCS) led the International Commission on Radiological Protection (ICRP) in 2011 to recommended the first ever single dose threshold for DCS independent of dose rate (ICRP 2012).
Yet, some uncertainties remain (e.g. as to plausibility of the assumed lack of dose protraction effects, the existence or otherwise of dose threshold) (Hamada and Fujimichi 2014;Hamada et al. 2014), necessitating a better understanding.
Regarding the impact of dose protraction, the latest systematic review and meta-analysis of epidemiological studies indicated larger excess DCS risks per unit dose in descending order of low dose rate exposure > acute exposure > fractionated exposure (Little et al. 2023).Such complicated patterns for sparing and enhancing dose protraction effects have also been observed in our experimental studies (Hamada et al. 2021(Hamada et al. , 2022)).Namely, our integrative analysis of multiple prelesional endpoints for damage to the thoracic aorta in wild-type mice at 6 months after starting acute, intermittent or continuous irradiation with 5 Gy of photons revealed the overall magnitude of aortic damage in descending order of X-rays in 25 fractions > acute X-rays > acute c-rays > X-rays in 100 fractions ) chronic c-rays (Hamada et al. 2021).Limitations of our previous work, nevertheless, included lack of data for fractionated c-ray exposure.Considering our previous observations of the greater biological effectiveness of acute X-ray exposure than acute c-ray exposure (throughout this paper, the phrase 'biological effectiveness' is used to mean the magnitude of biological effects), as well as the greatest effectiveness of X-ray exposure in 25 fractions (Hamada et al. 2021), here we set out to assess changes after c-ray exposure in 25 fractions, and compare its biological effectiveness with five other irradiation regimens.

Mice, shipping and irradiation
This study is composed of 12 groups for six irradiation regimens, as depicted in Figure 1.For each regimen, male C57BL/6J (B6J) mice started being total body irradiated with 5 Gy of photons, or sham (0 Gy)-irradiated for controls, at 8 weeks of age.For five regimens (all except the 'c-rays in 25 fractions' regimen), mice were shipped and irradiated with X-rays (260 kVp and 4.5 mA with a 0.5 mm Al and 0.3 mm Cu filter, 0.5 Gy/min) delivered acutely as a single dose (in 10 min), intermittently in 25 daily fractions (0.2 Gy/fraction spread over 42 days) or in 100 daily fractions (0.05 Gy/fraction over 153 days), or 137 Cs c-rays continuously (over 153 days at <1.4 mGy/h), as described (Hamada et al. 2021).Dosimetry was performed with glass dosimeters, as described (Hoshi et al. 2000;Sasatani et al. 2023).For the 'c-rays in 25 fractions' regimen, 7-week-old mice were shipped by car from Jackson Laboratory Japan (Shiga, Japan) to Hiroshima University (Hiroshima, Japan), and acclimated for a week.Ten unanesthetized mice were placed in a pie cage just before irradiation and were irradiated with 5 Gy of 137 Cs c-rays (0.5 Gy/min) intermittently in 25 daily fractions (0.2 Gy/fraction over 42 days).All control mice were sham-irradiated in parallel with the test mice, such that mice placed in a pie cage were put into an irradiator for the same time needed to complete irradiation, but were not actually irradiated.
All animal experiments were approved by the Animal Research and Ethics Committee of CRIEPI (approval number 2016-08) and the Institutional Animal Care and Use Committee of Hiroshima University (approval number A16-139), and carried out in compliance with the Japanese guidelines of animal care.

Echocardiography and tissue sampling
At 34 weeks of age (i.e. at 6 months after starting irradiation), mice were weighed, anesthetized with isoflurane, and subjected to motion/movement-mode (M-mode) echocardiography with a Toshiba Nemio MX SSA-590A ultrasound scanner and a Toshiba PLM-1202S ultrasound probe (representative images shown in Figure S1).To evaluate left ventricular function, various echocardiographic indices presented in Figure S2 were measured or calculated, as described (Hamada et al. 2021).Following echocardiography, mice were perfused transcardially with phosphatebuffered saline, and underwent tissue sampling.Heart and kidneys were weighed at tissue sampling (Figure S3).For each mouse, right and left kidneys were weighed together, and the mean was used as kidney weight.
Throughout the observation period of six months (i.e. at age 8-34 weeks) following the 'c-rays in 25 fractions' regimen, no mice died or appeared moribund, and there was no significant decrease in body weight (Figure S4).

Analysis of the aorta
The cranial half of the descending thoracic aorta was opened longitudinally, fixed and carbon coated, followed by fieldemission scanning electron microscopy (FE-SEM) to assess Experimental timelines.The present study consists of 12 groups of male C57BL/6J (B6J) mice for six irradiation regimens.Mice were shipped by car from Shiga to Hiroshima or from Kanagawa to Tokyo, and by car and air from Tokyo to Hiroshima, all in Japan.Mice were irradiated with 0 or 5 Gy of X-rays or c-rays, either acutely (in 10 min), intermittently (in 25 fractions over 42 days, in 100 fractions over 153 days) or continuously (over 153 days).Of six irradiation regimens, experiments with the 'c-rays in 25 fractions' regimen as highlighted in purple were newly conducted.For details, see the main text and a previous paper (Hamada et al. 2021).morphological changes, as described (Hamada et al. 2020).The surface of the normal aortic endothelium possesses vertical waves more frequently than horizontal waves (Figure 2(A)), and the crests in such vertical waves were counted in each of the seven fields/mouse (each field corresponds to the entire area of the image taken at 300 x magnification, 1 crest/field corresponding to $7.5 crests/mm 2 ).Each mouse was judged positive if there are any areas of 'detachment' (the detached area in the endothelium with sizes of a few tens of microns in the longer axis, Figure 2(B)) or 'large detachment' (that of the order of 100 lm, Figure 2(C)).Positivity was assessed in each group regardless of the number of such areas in each mouse: this was also the case for rolling leukocytes (Figure S6).
The caudal half of the descending thoracic aorta was embedded and snap-frozen, and the 5 lm thick transverse cryosections were mounted onto glass slides for staining, as described (Hamada et al. 2020).Dual immunofluorescence for cluster of differentiation 31 (CD31) stained green and one of the seven other markers (eNOS, VE-cadherin, TNFa, CD68, F4/80, CD3, TGF-b1) stained red, with cell nuclei counterstained with 4 0 ,6-diamidino-2-phenylindole (DAPI), was conducted and quantified, as described (Hamada et al. 2021).Table S1 lists antibodies used in immunofluorescence staining.For histochemistry, Masson's trichrome staining (where aniline blue stains collagen fibers in the tunica media blue) and Oil Red O staining (to stain neutral lipids in the atherosclerotic plaques red) were conducted and quantified, as described (Hamada et al. 2021).The intima-media thickness (IMT) was measured from immunofluorescence images and from histochemistry images.

Statistical analysis
The present paper adds new data for two (Figures 3, 5, and 6, S4-S7 and S9) or four (Figures S2,S3 and S10) groups to the previous paper (Hamada et al. 2021), and statistical analyses were reperformed for all the data in 12 groups using R statistical software (version 4.2.2,R Foundation, https:// www.r-project.org/accessed on 10 July 2023).A p-value (after Bonferroni corrections for pairwise comparisons using the t-test) of <0.05 was considered significant (p < 0.001 presented as ÃÃ , 0.001 p < 0.05 as Ã ), 0.05 p < 0.1 as marginally significant (presented as #) and p ! 0.1 as nonsignificant (presented as ns).In figures for six pairs of intraregimen comparisons (e.g.irradiated vs. sham-irradiated groups in each of six irradiation regimens), signs (i.e.asterisks, pounds and ns) were shown in black.In figures for 15 pairs of inter-regimen comparisons of such intra-regimen differences (i.e. the degree of differences in irradiated and sham-irradiated groups between the two regimens), asterisks and pounds were shown in blue, whereas ns signs were omitted for clarity.p-values determined by the one-way analysis of variance (ANOVA) using the F-test of homogeneity among the group means, a two-sample (Welch's) t-test for the null hypothesis of equal means, Wilcoxon rank sum test, Fisher's exact test, Wald test (logistic regression), and Kolmogorov-Smirnov goodness-of-fit test are presented as p.The statistical test used is described in figure legends or table footnotes.Each data was obtained from 8-10 mice and is shown in figures as means and standard deviations, unless otherwise specified.Statistical comparisons made for 21 pairs (six pairs of intra-regimen comparisons and 15 pairs of inter-regimen comparisons) in each endpoint are outlined in figure legends and are not repeated elsewhere.
To compare biological effectiveness of each irradiation regimen, the integrative analysis was performed with two approaches.One approach was to convert different significance levels (indicated as ÃÃ , Ã , # and ns) to discrete scores (±1.5, ±1, ±0.5 and 0, respectively) for each of 15 interregimen comparisons among six irradiation regimens for each endpoint (Tables S3 and S5).For each inter-regimen comparison (e.g.acute X-rays vs acute c-rays), a difference in effectiveness was judged based on mean scores from multiple endpoints according to the criteria: ) (mean score !0.5), > (0.2 mean score <0.5), !(0.1 mean score <0.2), and ¼ (0.1 > mean score).The other approach was Kolmogorov-Smirnov goodness-of-fit test (Tables S4 and  S6).For comparison groups 'A' (e.g.acute X-rays) and 'B' (e.g.acute c-rays), alternative hypotheses 'jAj > jBj' and 'jAj < jBj' were tested for multiple endpoints, vis-a-vis the null hypothesis that the absolute values of the radiation effects between the given comparison groups are equal for all endpoints (jAj ¼ jBj).For each inter-regimen comparison, a difference in effectiveness was judged based on p values (of one-sided Kolmogorov-Smirnov goodness-of-fit test for the null hypothesis that the distribution of the p values from the individual tests is the standard uniform, which is true with no difference in radiation effects for all endpoints between the given comparison groups) according to the criteria for jAj > jBj: ) (p < 1 Â 10 À5 ), > (1 Â 10 À5 p < 1 Â 10 À1 ), !(1 Â 10 À1 p < 3 Â 10 À1 ) and ¼ (p ! 3 Â 10 À1 ).For both approaches, the overall difference among six irradiation regimens were judged based on differences in each of 15 interregimen comparisons.Such integrative analyses were made, first for all aortic endpoints tested (Tables S3 and S4), and then for only endpoints that showed significant difference between irradiated and sham-irradiated groups at least in one irradiation regimen (Tables S5 and S6).

Results
In this study, there are 12 groups for six irradiation regimens, and 31 prelesional endpoints consist of 8 for echocardiography, 5 for body or organ weight, and 18 for the aorta (4 for FE-SEM, 12 for immunofluorescence, and 2 for Masson's trichrome staining).Irradiated mice exhibited changes in 23 out of 31 endpoints (15 increased, 8 decreased) at least in one irradiation regimen, compared with sham-irradiated mice (Table S2).Among them, echocardiography and body or organ weight demonstrated slight changes in 6 of 13 endpoints (1 increased, 5 decreased, with significant differences only in two endpoints) in 3 of 6 irradiation regimens, but with significant differences only in two endpoints in one irradiation regimen (Figures S2 and S3).On the contrary, the aorta showed significant differences in 17 of 18 endpoints in all irradiation regimens (Figures S5-S7 and S9-S11), and is therefore a focus of the following sections.The first subsection describes the results for the 'c-rays in 25 fractions' regimen, and the second subsection describes the results of the integrative analysis for aortic endpoints in six irradiation regimens.

Responses following exposure to c-rays in 25 fractions
Irradiation reduced waviness in the aortic endothelium (Figure 3(A)) due to various morphological changes such as flattening, derangement and cobblestone formation (Figure 2(D-F)), and resulted in detachment and large detachment of the aortic surface (Figures 2(B,C) and 3(B,C)).In vascular endothelial cells (VECs) following irradiation, CD31 negativity (indicative of VEC or CD31 loss), DAPI negativity (indicative of VEC loss) and subcellular fragments (indicating apoptosis) increased (Figures 5(A,B), S8 and S9A), whereas eNOS (a marker for vascular functionality) and VE-cadherin (a marker for adherens junctions) decreased (Figures 4(A,B) and 5(C,D)).These findings indicate that irradiation produces vascular damage through partial loss of the aortic endothelium behind which mechanisms include VEC apoptosis.Following irradiation, vascular smooth muscle cells (VSMCc) exhibited increases in TNF-a (a marker for proinflammation), CD68 and F4/80 (markers for macrophages), CD3 (a marker for T cells) and subcellular fragments (indicating apoptosis) (Figures 4(C-F) and 5(E-H), S8 and S9B), indicating that radiation-induced vascular damage causes inflammation.The aortic wall showed increases in TGF-b1 (a marker for profibrosis), alanine blue stain (a marker for collagen fiber) and IMT (determined by two approaches) (Figures 4(G), 5(I,J), and 6 and S11), indicating that irradiation induces fibrosis.Vascular damage, inflammation and fibrosis observed here have all been implicated in the early stages of atherosclerosis.Nevertheless, no aortae in irradiated or sham-irradiated mice were positive with Oil Red O (Figure S12), suggesting the absence of mature atherosclerotic lesions in the aorta of B6J mice.Overall, the responses following exposure to c-rays in 25 fractions were qualitatively very similar to those following other irradiation regimens, albeit to a different degree quantitatively as detailed in the next subsection.

Biological effectiveness of six irradiation regimens
To compare biological effectiveness of six irradiation regimens, the integrative analysis was performed with two approaches (one approach with different significance levels converted to discrete scores, the other approach with Kolmogorov-Smirnov goodness-of-fit test).Such integrative analyses suggested effectiveness in descending order of 'X-rays in 25 fractions > (or !) acute X-rays > c-rays in 25 fractions > acute c-rays > Xrays in 100 fractions > chronic c-rays' for 18 prelesional aortic endpoints (i.e.all endpoints tested) (Tables S3 and S4) and 'Xrays in 25 fractions > acute X-rays > c-rays in 25 fractions > acute c-rays > X-rays in 100 fractions ) chronic c-rays' for 16 prelesional aortic endpoints (i.e.exclusive of two endpoints without significant differences in any of the 15 pairs among six irradiation regimens) (Tables S5 and S6).

Discussion
Here we have demonstrated that dose protraction effects on aortic damage depend on radiation quality, and are not a simple function of dose rate and the number of fractions.As such, the following subsections focus first on radiation quality (X-rays vs c-rays) and then on dose protraction effects (sparing vs enhancing), while recognizing that this study could overcome only one of various limitations of the study design by adding the data for exposure to c-rays in 25 fractions: all other limitations would remain such as interregimen differences other than in irradiation (e.g. in shipping, feed, the beddings, and use of a turntable during irradiation or sham-irradiation), lack of exposure to c-rays in 100 fractions and chronic X-rays, inconsistent duration of irradiation (42 days for 25 fractions, 153 days for 100 fractions and chronic exposure), lack of acute exposure at age 14 and 30 weeks (when delivery of 25 fractions and chronic exposure completes, respectively), and lack of multiple dosepoints, which have previously been discussed at length (Hamada et al. 2021) and are not repeated here.

Radiation quality
Our previous work showed that acute X-rays are more effective than acute c-rays (Hamada et al. 2021).The present study extended this observation, by finding that such higher effectiveness of X-rays (260 kVp) vs 137 Cs c-rays (662 keV) can also be observed after exposure in 25 fractions and that there lies effectiveness of c-rays in 25 fractions between acute X-rays and acute c-rays (Tables S3-S6).The actual biological effectiveness per unit dose can be higher for Xrays than c-rays, because with the same entrance skin dose (i.e. 5 Gy in this study), depth dose (the dose at the organ of interest, e.g.aorta in this study) can be lower for X-rays than c-rays.These findings are in line with a body of in vitro evidence that lower energy photons possess higher effectiveness following acute exposure (NCRP 2018;Hamada et al. 2006), such that $200 kV X-rays are 1.5-2 fold more effective than 137 Cs c-rays for induction of chromosome aberrations (Schmid et al. 2002) and H2AX focus formation (Korns et al. 2023).On the contrary, such evidence is very limited in vivo, e.g. for tumor control (Baumann et al. 1999), to which this study could add for better understanding of potential differences in biological effectiveness with energy of photons.Not only direct effects to the aorta but also effects from various other organs/tissues (including the heart, kidneys, and other potential target organs/tissues for radiation effects on the circulatory system) should contribute to a series of aortic changes observed following total body irradiation, and more studies are needed to elucidate how such effectiveness differs with energy of photons.Taken together, it is worth noting that acute exposure to X-rays, but not that to c-rays, increased TGF-b1 and F4/80, with great differences in these two endpoints between X-ray and c-rays when delivered in 25 fractions (Figure S7G and S7I), warranting further studies.

Dose protraction effects
We showed that X-rays in 100 fractions and chronic c-rays are less effective than acute exposure (Tables S3-S6).Such sparing dose protraction effects follow a tenet that biological effectiveness decreases as dose rate decreases (R€ uhm et al. 2015), to which some recent epidemiological studies on DCS mortality lends support (Sasaki et al. 2020;Azizova et al. 2023).Chronic exposure to 137 Cs c-rays for 400 days at 20 mGy/day (c.f., for 153 days at <34 mGy/day in our chronic exposure regimen) has also been shown to induce various non-neoplastic lesions in the adrenal, liver and ovaries (Braga-Tanaka et al. 2018), but it remains unclear whether exposure at lower dose rate is less effective.On the other hand, our previous work also found that X-rays in 25 fractions are more effective than acute X-rays (Hamada et al. 2021).The present study extended this observation, by finding that such higher effectiveness of exposure in 25 fractions vis-a-vis acute exposure can also be observed after c-rays (Tables S3-S6).Such potential enhancing (inverse) dose protraction effects are supported by the similar observations in human umbilical vein endothelial cells in vitro (Cervelli et al. 2014), in kidneys of B6J mice and DCS mortality of B6CF 1 mice in vivo (Seol et al. 2012;Hoel and Carnes 2017;Tran and Little 2017), and in the Canadian tuberculosis fluoroscopy cohort (Zablotska et al. 2014).Its mechanistic underpinnings are unclear (Tapio et al. 2021), but there are several possibilities, e.g. that biological effectiveness increases with decreasing dose (i.e.inverse dose effects), and that interaction of newly formed damage with existing damage leads to production of greater damage.Mounting evidence in Japanese atomic bomb survivors (Little et al. 2020;Little and Hamada 2022) and in other cohorts (Little et al. 2023) for higher DCS risk per unit dose at lower dose is supportive of the former possibility.
The current ICRP System of Radiological Protection recommends a single dose threshold for DCS independent of the rate of dose delivery assuming that there would be no dose protraction effect (ICRP 2012), but implications of complicated patterns of dose protraction effects for radiation protection necessitate further extensive discussion (Hamada 2023).The observation of both enhancing and sparing dose protraction effects also beckons a question of whether there is any borderline dose fractionation regimen, above which enhancing dose protraction effects would occur, but below which sparing dose protraction effects would occur.The observation that chronic c-rays are less effective than X-rays in 100 fractions in spite of the same duration of exposure (i.e.153 days) for both regimens (Tables S3-S6) also invites another question as to whether there is any borderline dose fractionation regimen that is iso-effective with chronic exposure (e.g.>100 fractions as suggested by this study).
Inter-regimen comparisons of 18 aortic endpoints (Table S3) suggest that two endpoints (aortic VECs and VSMCs with subcellular fragments, indicative of apoptosis) with mean scores >1.2 contributed most greatly to inter-regimen differences, followed by three endpoints (stained intensity per unit aortic wall area, TGF-b1 and TNF-a) with mean scores 0.8-0.9.In contrast, CD3 (mean score 0) exhibited post-irradiation changes all after six irradiation regimens, but showed no inter-regimen difference (Figure S7H, Table S3).It is also intriguing to note that in contrast to the overall inter-regimen differences, two endpoints (CD68 and TGF-b1) demonstrated opposite changes in two irradiation regimens, and this was also the case for one endpoint (aortic VECs with subcellular fragments) in one irradiation regimen (Tables S3 and S6).These observations merit further studies.

Conclusions
The present study showed that irradiation causes vascular damage, inflammation and fibrosis in the aorta of wild-type mice.These prelesional changes have all been implicated in the early stages of atherosclerosis, but no mice actually developed mature atherosclerotic lesions.The integrative analysis for a series of prelesional aortic endpoints revealed that dose protraction effects on aortic damage depend on radiation quality, and are not a simple function of dose rate and the number of fractions.This raises a question of whether other large arteries implicated in radiation DCS pathogenesis also exhibit such complicated patterns for sparing and enhancing dose protraction effects, and further studies with the carotid artery are thence under way.

Figure 1 .
Figure1.Experimental timelines.The present study consists of 12 groups of male C57BL/6J (B6J) mice for six irradiation regimens.Mice were shipped by car from Shiga to Hiroshima or from Kanagawa to Tokyo, and by car and air from Tokyo to Hiroshima, all in Japan.Mice were irradiated with 0 or 5 Gy of X-rays or c-rays, either acutely (in 10 min), intermittently (in 25 fractions over 42 days, in 100 fractions over 153 days) or continuously (over 153 days).Of six irradiation regimens, experiments with the 'c-rays in 25 fractions' regimen as highlighted in purple were newly conducted.For details, see the main text and a previous paper(Hamada et al. 2021).

Figure 2 .
Figure 2. Representative FE-SEM images for morphology of aortic endothelium.(A) Normal endothelium.(B) Endothelium with detachment.(C) Endothelium with large detachment.(D) Flattened endothelium.(E) Deranged endothelium.(F) Cobblestone (or snap pea)-shaped endothelium.All images were taken at 6 months after starting irradiation with 0 Gy (A) or 5 Gy (B-F) of c-rays in 25 fractions.(A-C) Boxed areas in the upper panels are shown at higher magnification in the lower panels.Scale bars are as indicated.For quantitative data, see Figures 3, S5 and S6.

Figure 3 .
Figure 3. Quantitative analysis of morphological alterations in the aortic endothelium for (A) the number of crests/field, (B) percentage of mice with detachment, and (C) percentage of mice with large detachment, following exposure to c-rays in 25 fractions (10 mice/group analyzed, Welch's t-test, Fisher's exact test).ÃÃ , p < 0.001.Ã , 0.001 p < 0.05.For representative FE-SEM images, see Figure 2.For comparison with other irradiation regimens, see Figure S5.

Figure 4 .
Figure 4. Representative immunofluorescence images for molecular changes in the aorta.The aorta underwent dual immunofluorescence of CD31 (green) with (A) eNOS, (B) VE-cadherin, (C) TNF-a, (D) CD68, (E) F4/80, (F) CD3 or (G) TGF-b1 (red), with cell nuclei counterstained with DAPI (blue).All images were taken at 6 months after starting irradiation with 0 Gy (upper panels) or 5 Gy (lower panels) of c-rays in 25 fractions.Boxed areas in the left panels (tiled images) are shown at higher magnification in the right panels.Scale bars are as indicated.For quantitative data, see Figures 5 and S7.

Figure 6 .
Figure 6.Masson's trichrome staining.(A) Representative images for masson's trichrome staining of the aorta.All images were taken at 6 months after starting irradiation with 0 Gy (upper panels) or 5 Gy (lower panels) of c-rays in 25 fractions.Boxed areas in the left panels (tiled images) are shown at higher magnification in the right panels.Scale bars are as indicated.Quantitative analysis of fibrotic alterations in the aorta for (B) intensity of alanine blue, and (C) IMT (10 mice/group analyzed, Welch's t-test or Wilcoxon rank sum test).AU, arbitrary unit.ÃÃ , p < 0.001.Ã , 0.001 p < 0.05.For comparison with other irradiation regimens, see Figure S10.