Effects of Continuous In Utero Low- and Medium-Dose-Rate Gamma-Ray Exposure on Fetal Germ Cells

Epidemiological studies have shown that in utero radiation exposure from atomic bombs, and from diagnostic and therapeutic procedures, results in a variety of health effects, including growth retardation, microcephaly, mental retardation (1 ) and increased cancer incidences (2, 3 ). However, it has been argued that the epidemiological studies on in utero exposures in humans contain uncertainties, due to limited information, uncertainties in dose estimation and questionable biological plausibility (4 ). Verreet et al. (5 ) similarly noted the limitations of epidemiological studies in humans and the necessity of good animal studies in their investigation of brain defects after prenatal radiation exposure. Animal studies have shown that the effects of acute high-dose irradiation in utero include growth retardation and congenital anomalies, including those of the central nervous system, and cognitive and behavioral abnormalities (6, 7, 8 ), as well as increased cancer incidence (9 ) in surviving fetuses. In several published animal studies (10 ), including our previously reported work (11 ), chronic exposures were used. Although these studies showed that chronic exposure can also induce various effects, they are usually undetectable in the low-dose-rate (LDR) range (<0.1 mGy/min or <144 mGy/day) (12 ). In our recently reported study, chronic in utero irradiation for the entire gestation period (18 days) at medium dose rates (MDRs) (200 and 400 mGy/day) did not induce congenital anomalies that are often observed in the fetuses exposed to 2 Gy in utero at high dose rate (HDR) on gestation day (GD) 11. Medium-dose-rate radiation induced a mild degree of general growth retardation as evidenced by low body weight from birth (average body weight at 10 weeks of age was 89–103% of the nonirradiated controls) (11 ). The effects on the gonads or germ cells were striking at 10 weeks of age with the testes and ovaries weighing approximately 11–22% of the nonirradiated controls.

Research on the effects of radiation to the gonads in both humans and animals show that germ cells are highly radiosensitive (13, 14 ). It has been shown that germ cells have very unique developmental processes (15–17 ); thus, the radiosensitivity of embryonic germ cells is of great interest. Somatic cells, such as Sertoli cells and Leydig cells, on the other hand, are more resistant to the effects of radiation than germ cells (18–22 ). It is known that characteristics of the gonadal somatic cells, such as gene expression and expression of sexual hormone receptors, change from fetal life through sexual maturity (23–25 ). Therefore, the radiosensitivity of the cells in the gonads, besides differences in their cell origin, may differ also before and after birth. Since most of the above-mentioned studies were conducted using HDR and/or high-dose irradiation, studies on chronic irradiation effects, especially at LDR, are very important from the perspective of radiation protection (13, 14 ). In general, the developing germ cells are considered to be highly radiosensitive. Most of the animal studies on germ cells have been done on postnatal germ cells, but in a few of these studies (26–30 ), the effects of in utero exposure on the fetal gonads and germ cells have been examined. These studies showed that mice oocytes (LD50: ∼0.15 Gy) are more radiosensitive than that of rats' (LD50:1 Gy) (31 ) suggesting that mice are more suitable animals for observing the effects of radiation and for exploring the mechanisms at low dose and LDR.

The use of lethal doses at acute high doses in studies of the animal germ cell precludes the extrapolation of the results since the doses do not represent the human environmental exposures that are typically chronic low-dose (dose-rate) exposures (32 ). Although the biological effects of low-dose and LDR radiation are issues of great interest, epidemiological studies of humans with long-term exposure to low-dose radiation are full of uncertainties (33 ). To elucidate these effects and their mechanisms, animal experiments in which radiation dose and other conditions can be strictly controlled are considered essential.

In this study, we have investigated the development of gonads in GD 18 fetuses that received MDR (200 and 400 mGy/day) gamma-ray irradiation in utero for the entire gestation period by examining the histomorphological changes and cell counts of germ cells in both sexes and Sertoli cell counts in males. Using these same parameters, we also analyzed the effects of the in utero chronic exposure at a LDR of 20 mGy/day wherein no gross effects were detected in our previously reported study (11 ).

Specific-pathogen-free (SPF) C57BL/6JJcl female mice and C3H/ HeNJcl male mice, purchased from CLEA Japan, Inc. (Tokyo, Japan), were housed individually in plastic cages (218 × 320 × 133 mm). For breeding purposes, virgin female C57BL/6JJcl mice (10–30 weeks of age), were transferred into the cage of C3H/HeNJcl male mice (1:1) in the afternoon and were allowed to mate overnight. The following morning, females with confirmed vaginal plugs were considered pregnant at GD 0 and were then transferred into individual cages and randomly assigned to nonirradiated control or irradiated dose groups. All mice were inspected once a day during the irradiation period as described elsewhere (34 ).

We elected to use B6C3F1 mice for this study since our previously reported studies (11, 34, 35 ) also used the same strain for comparison purposes. Experiments are also underway to study the late effects, such as life span and neoplasm incidence, of chronic exposure to gamma rays in utero under similar conditions using the same strain.

Pregnant dams were randomly assigned to groups receiving whole-body gamma-ray irradiation continuously for 22 h each day from GD 0 to 18 at a LDR (20 mGy/day, total dose = 360 mGy) or at MDRs (200 mGy/day, total dose = 3,600 mGy or 400 mGy/day, total dose = 7,200 mGy) as described elsewhere (11 ). The remaining 2 h were used for health monitoring of the mice and husbandry procedures, such as cleaning the rooms, cage changes with fresh bedding and providing a fresh supply of food and water.

The LDR of 20 mGy/day was selected based on our previous study wherein we reported life span shortening in mice exposed continuously for 400 days (35 ). The MDRs of 200 and 400 mGy/day were based on our other previous study where we reported hypoplasia of the gonads in mice exposed in utero (11 ).

On GD 18, a total of 29 dams were euthanized with an overdose of sevoflurane (SevoFlo®; DS Pharma Animal Health Co. Ltd., Osaka, Japan) and a total of 247 fetuses were collected via cesarean section (Table 1). Each fetus was assigned a unique identification number based on its location in the uterus following standard operating procedures and was humanely euthanized by submersion in ice-cold phosphate-buffered saline (PBS, pH 7.4; TaKaRa Bio Inc., Shiga, Japan). To visualize the implantation sites for counting, the empty uteri were clarified by immersing them in 0.2% NaOH (36 ).

Fetal gonads were collected and fixed in 10% neutral-buffered formalin. Serial paraffin-embedded sections (4 µm thick) were prepared for hematoxylin and eosin (H&E) staining and for immunohistochemistry. Sections with the largest cross-sectional area were selected for use.

Sections of the gonads were autoclaved in antigen retrieval solution (Agilent Technologies, Cheadle, UK) at 121°C for 10 min, followed by incubation with a blocking reagent (Agilent Technologies) at room temperature for 30 min. The sections were then sequentially incubated with anti-MVH antibody (Mouse Vasa Homolog or DDX4; Abcam, Cambridge, UK) and SOX9 (Abcam) at 4°C overnight, EnVision+ System-HRP anti-rabbit (Agilent Technologies) at room temperature for 30 min, 3,3-diaminobenzidine (DAB) solution of Liquid DAB+ Substrate Chromogen System (Agilent Technologies), and finally stained with hematoxylin.

Images of the stained slides of the fetal gonads were obtained using an Olympus BX51 microscope and analyzed using the Olympus cellSens^™ imaging software (Tokyo, Japan). The numbers of MVH- and SOX9-positive cells were scored, and the total gonadal and the seminiferous tubule areas measured (37, 38 ).

All statistical analyses were carried out using R version 3.6.2 (39 ) with packages lme4 and multcomp. Generalized linear models (GLMs) or generalized linear mixed models (GLMMs) were used to correct for litter effect (40, 41 ). We treated the irradiation group as a fixed effect for all GLM or GLMM analyses. GLMs were used for the analyzing the average number of the fetuses and implantation sites. GLMMs were used for the analyses of body weights of fetuses, gonadal area, seminiferous tubule area, and stromal area with litter size as a random effect. For analyses of the number of MVH-positive cells in each sex and SOX9-positive cells in the testis of males, GLMMs were used with litter as a random effect and log-transformed seminiferous tubule area for males or ovary area for females included as an off-set term. Discrete data (average number of live fetuses and implantation sites, numbers of MVH⁺ cells and SOX9⁺ cells) were modeled with a Poisson distribution (using a log-link function). Gonadal, seminiferous tubule and stromal areas were modeled with a Gaussian distribution (using an identity function). Post hoc comparisons for the number of live fetuses, implantation sites, body weights of fetuses, gonadal areas and seminiferous tubule areas of male testes were performed using Dunnett's test. Effects were considered statistically significant at P < 0.05.

Technical problems (inadequate/insufficient sample, i.e., due to the small size and immunostaining failures) prevented us from analyzing all the collected gonads for gonadal area, seminiferous tubule area and germ cell numbers.

Pregnant mice were irradiated from GD 0 to 18 at a LDR of 20 mGy/day or at the MDRs of 200 or 400 mGy/day. The fetuses and the uteri were collected and examined on GD 18. Table 1 shows the number of the fetuses and implantation sites/dam, and the average body weights of the fetuses. There was no significant difference in the number of the fetuses or the implantation sites between the nonirradiated control and each irradiated group. The average body weights of the fetuses in both sexes from the LDR group were not significantly different from the nonirradiated control, whereas those from the MDR groups were significantly lower, by 76–89%, than the nonirradiated controls. These results are consistent with those of our previously reported study (11 ).

The gonads of the fetuses were histomorphologically examined and quantitatively analyzed. Figures 1 and 2 show the representative sections of the testes and the ovaries, respectively. The gonadal area was measured and the numbers of the germ cells (MVH⁺) and the Sertoli cells (SOX9⁺) were recorded. The seminiferous tubule areas were also measured. These values were analyzed and are shown in Figs. 3 and 4.

The cross-sectional areas of both testes and ovaries from the MDR groups was 56–75% that of the nonirradiated controls (Figs. 3A and 4A). Smaller testes from the MDR groups also had significantly decreased (45–55%; P < 0.001) seminiferous tubule areas (Fig. 3A), whereas the interstitial areas (calculated by subtracting the seminiferous tubule area from the gonadal area) did not change significantly. No histomorphological abnormalities (e.g., hyperplasia) were observed in the stromal Leydig cells (Fig. 1). The number of the MVH-positive germ cells in the seminiferous tubule area and the ovaries from the MDR groups were markedly decreased (P < 0.001) or almost absent (Figs. 3B and 4B). A slight but significant increase (P < 0.001 and P < 0.01) in the number of Sertoli cells was observed in the seminiferous tubule area in both 200 and 400 mGy/day MDR groups, respectively (Fig. 3C). There was no significant difference in the number of Sertoli cells/ seminiferous tubule between the nonirradiated control and MDR-irradiated groups (data not shown).

Although no histomorphological changes were observed in the gonads of fetuses irradiated at LDR, a non-significant decrease (P = 0.34 for males and P = 0.10 for females) in the number of germ cells was observed (Figs. 1 and 2).

The results of the current study indicate that chronic in utero γ-ray irradiation at medium dose rates (200 and 400 mGy/day) during the entire gestation period induces very significant effects on the development of gonads, specifically the obliteration of most of the germ cells. The significant decrease in seminiferous tubule area with no significant change in the stromal area in the MDR groups indicate that the reduction in testicular area is mainly due to hypoplasia of the testicular parenchyma (seminiferous tubules) (Figs. 1 and 3A).

Radiosensitivity of germ cells is influenced by sex, age, radiation dose and gametogenic stage (13, 14 ). Studies on the effects of radiation on the gonads and the germ cells, mostly postnatal exposure at high doses or high dose rates, show that they are highly radiosensitive, particularly female germ cells (42–44 ). Since the effects on the testes (or the male germ cells) and ovaries (or the female germ cells) observed in this study appear to be similar in magnitude, we hypothesize that radiation affects the primordial germ cells, prior to differentiation into male and female germ cells at approximately GD 12.5 (45 ). There is, however, a need to identify or pinpoint the exact stage of embryonic development wherein the germ cells are most sensitive to radiation, as there have been only a few reports on the effects of in utero radiation exposure on the gonads (28–30 ).

Sertoli cells are thought to be relatively more resistant to radiation than germ cells (46 ). In this study, the number of Sertoli cells in the seminiferous tubule area were not decreased but rather significantly increased in the MDR groups (Fig 3C). It has been shown that the characteristics of Sertoli cells, such as gene expression and expression of sexual hormone receptors, change from fetal life through sexual maturity (25 ), which may, in turn, lead to variations in radiosensitivity. Chronic MDR irradiation may be a proliferative factor for Sertoli cells, but there is currently no evidence to support a hormesis effect in the male genital system (13 ). Poorly differentiated or underdeveloped seminiferous tubules, however, may also create a semblance of an increase in the number of Sertoli cells in the MDR groups. The effect of MDR on the number of Sertoli cells needs to be clarified further. Additionally, it is also known that the somatic components of the male gonads, especially Sertoli cells, interact with the germ cells during development (17 ). Therefore, more innovative investigative methods including functional analyses would be necessary to explore these.

No obvious histological changes were observed in the testicular stroma in the current study. In both the MDR-irradiated fetuses at GD 18 in this study and at 10 weeks of age in our previous study, the testes were smaller, in size and weight, than the nonirradiated controls, but with an ample number of Leydig cells in the interstitium. These hormone-producing Leydig cells appear to be relatively resistant to chronic MDR irradiation, consistent with previous studies citing their relatively high radioresistance (46–48 ). In adult animals, it is has been reported that radiation causes dysfunctional Leydig cell depletion (20 ) or temporary hyperplasia (22 ). As with Sertoli cells, Leydig cell characteristics change from fetal and adult stages (23, 24 ), and further work is necessary to determine how radiation can affect each developmental stage of Leydig cells (e.g., their proportions).

It has long been established that irradiation of the ovaries accelerates oocyte depletion, with consequent declined hormone production and premature menopause (41, 46, 48 ). Similar studies on the effects of in utero exposure to chemicals (49–52 ), cigarette smoke (53 ), anti-cancer drug (54 ) and other drugs (55 ) have demonstrated disruption of ovarian follicles and development by altering the gene and protein expressions in the germ cells and/or somatic cells, thereby affecting reproductive function, including the estrus cycle. In utero exposures of sheep via the mother to low doses of environmental chemicals (56 ) and of humans to cigarette smoke (57 ) have led to similar effects, suggesting a common mechanism regardless of species. In addition, in utero exposures to some chemicals (50, 53, 58 ) and drugs (55 ) have induced reproductive deficits and alterations in gene expression in the offspring, indicating transgenerational effects. In our previously published study, the major hormone-producing structures of the ovary, specifically the ovarian follicles and corpora lutea, were very few or nonexistent in 10-week-old mice irradiated at MDR in utero (11 ). As with the developing testes, the numerous interactions between the germ and somatic cell lineages are important for the differentiation of somatic tissue in the ovary (17 ).

Gender differences in radiosensitivity in both humans (Japanese atomic bomb survivors, Chernobyl cohort, occupational exposure cohorts) and animals and have been recently reviewed by Narendran et al. (59 ). It has been posited that for late effects, such as altered tumor spectra and shortened life spans in mice irradiated in utero (60 ), hormonal factors, such as estrogen or prolactin, may contribute to gender differences.

The decreases in the testicular and seminiferous tubule areas of the testes (Fig. 3A) and ovary areas (Fig. 4A), as well as in the number of the germ cells observed in the fetuses before birth (at GD 18) in this study (Figs. 3B and 4B) are consistent with the findings in the mice after maturity (at 10 weeks of age) in our previous study (11 ). These findings confirm that the hypoplasia of the gonads and the depletion of the germ cells observed at 10 weeks of age in our previous study (11 ) were present around the time of birth (GD 18) and that damage is permanent and irreversible rather than transient.

The changes we report here represent the primary effects of radiation exposure on the gonads during the early stages of development. The gonads are also strongly influenced by the pituitary gland (pituitary-gonadal axis), especially after sexual maturation. As mentioned previously, most in utero radiation exposures (acute high doses), reported to date, focus on developmental anomalies, embryo resorption, fetal death, growth retardation and cancer incidence as end points. Our results elucidate the need for further investigation into the long-term and late effects of chronic radiation exposures in utero at low and medium dose rates using reproductive performance (mating performance, pregnancy rates, etc.), carcinogenesis and life span as end points.

While drastic changes were observed in MDR-irradiated mice, a slight, but not significant, reduction in the number of germ cells was observed in LDR-irradiated mice. This suggests that it is possible to observe effects at lower doses, particularly in exposures during periods of high sensitivity, as described in previously published studies. Another possibility is that at LDR or chronic radiation exposures, the mechanism by which it affects the germ cells may differ from that of acute or HDR exposures.

Since the decrease in the number of the germ cells in the testes and the ovaries in the fetuses exposed in utero to MDR radiation was quite extreme, we consider it to be one of the most sensitive indicators of in utero irradiation. Other than the non-significant decrease in the number of germ cells in the fetuses exposed to LDR (20 mGy/day) radiation, we also did not observe any obvious differences in Sertoli or Leydig cells. Intrinsic differences in gametogenesis between males and females, such as differences in the duration of meiosis, suggest gender differences in response to low-dose chronic irradiation, and warrants further investigation (32 ).

Further investigation on the effects of chronic LDR radiation exposure in utero is necessary, and we hope that our above-mentioned ongoing studies on the late effects (reproductive capability, tumorigenesis and life span) may further clarify these effects, as these reflect the in vivoamplified consequences of the effects of radiation on the gonads and the germ cells.

We are indebted to the technical and animal care staff of IES for their support during the conduct of the study. This work was performed under contract with the government of Aomori Prefecture, Japan.