Temporal variation of excess mortality rate from solid tumors in mice irradiated at various ages with gamma rays

Reasons why the idea that radiation exposures induce cancer needs to be revisited

PURPOSE

It has long been thought that the carcinogenic effect of radiation resulted from the induction of oncogenic mutations which then led to an increase in the proportion of cancer-bearing individuals. However, even as early as the 1960s, there were indications that the carcinogenic effect of radiation might result from the induction of an earlier onset of cancer. Recently, the former notion was challenged by its inability to explain time-dependent decline of the relative risk following an exposure to radiation, and a parallel shift of mouse survival curves toward younger ages following an exposure to radiation. The two observations are clearly understood if it is assumed only that a radiation exposure causes an earlier onset of spontaneously occurring cancers.

METHOD

In the present study, a critical review was conducted which examined papers that showed dose responses which apparently supported the mutation induction theory of radiation carcinogenesis.

RESULTS

It was found that there were two types of misleading experimental designs: one consisted of studies in which observations were prematurely terminated, and which consequently hid a complete story of radiation carcinogenesis. The other set of papers used age adjustments which were derived from the idea that the life shortening effect of radiation needs to be compensated for since tumor mortality becomes higher among older subjects. This type of adjustment appeared reasonable but was found actually to be a different form of description on an earlier onset of cancer following radiation exposures.

CONCLUSION

In mouse experiments, radiation exposures did not lead to the induction of a large increase in the proportion of tumor deaths when life-long observations were made. Human epidemiologic data are also in line with the earlier onset hypothesis of radiation action. It should be cautioned, however, that the earlier onset model applies only to malignancies whose mortality increases rapidly with the increase of age and does not apply to diseases of short latency such as childhood leukemia and thyroid cancers.

Molecular and cellular basis of the dose-rate-dependent adverse effects of radiation exposure in animal models. Part II: Hematopoietic system, lung and liver

While epidemiological data have greatly contributed to the estimation of the dose and dose-rate effectiveness factor (DDREF) for human populations, studies using animal models have made significant contributions to provide quantitative data with mechanistic insights. The current article aims at compiling the animal studies, specific to rodents, with reference to the dose-rate effects of cancer development. This review focuses specifically on the results that explain the biological mechanisms underlying dose-rate effects and their potential involvement in radiation-induced carcinogenic processes. Since the adverse outcome pathway (AOP) concept together with the key events holds promise for improving the estimation of radiation risk at low doses and low dose-rates, the review intends to scrutinize dose-rate dependency of the key events in animal models and to consider novel key events involved in the dose-rate effects, which enables identification of important underlying mechanisms for linking animal experimental and human epidemiological studies in a unified manner.

Effects of irradiation on cumulative mortality in mice: shifting toward a younger age of death

Recently, the question of whether cancer risk is only accelerated but not increased by radiation exposure has been raised. To explore this matter, we analyzed whether the cumulative mortality of irradiated mice could be explained by x-axis (age) shifted cumulative mortality of nonirradiated mice. We reanalyzed publicly available data on observed cumulative mortality or prevalence in irradiated female B6C3F1 mice that lived their entire lifespan. The results showed that the irradiated curve was well matched to uniformly shifted nonirradiated curve for the cumulative mortality of all causes of death but not for the cumulative mortality of all solid tumors and prevalence of ovarian tumors as is. After adjusting lifetime mortalities, it was also well matched for all solid and ovarian tumors. The shifted days by irradiation were 71-116 days for all causes of death, 56-135 days for all solid tumors, and 41-140 days for ovarian tumors in the 1.9 Gy-irradiated group. The response was switched between irradiation at 35 and 105 days consistently for all the above indexes, supporting the hypothesis that radiation sensitivity differs between juvenile and adults. The shifted days of all causes of death showed a tendency of linear response to dose. This concept of shifting the age of death can be applied not only for all cause of death but also for mortality of all solid tumors after adjusting the magnitude. These findings contribute to the discussion on the application of the 'shifting age of death' concept to radiation protection.

Radiation-induced increases in cancer mortality result from an earlier onset of the disease in mice and atomic bomb survivors

PURPOSE

It has long been thought that the carcinogenic effect of radiation is due to the induction of oncogenic mutations, which means that a fraction of the irradiated individuals will be affected in a dose-dependent manner. This dogma was recently challenged because it was found that the model does not properly explain the life shortening effect of radiation which is seen as a parallel shift of mouse survival curves toward younger ages following an exposure to radiation. Specifically, according to the mutation induction theory, an irradiated mouse or human population evolves into two subpopulations with different mean lifespans, which would lead to a wider distribution of individual lifespans, and hence to a shallower slope in the survival curve, which is not what is observed. Instead, the parallel shift indicates that a large fraction of the irradiated mice are affected (but there are exceptions). Thus, it was thought important to pursue how the excess risk for cancer develops following an exposure to radiation.

METHOD

In the present study, cancer mortality data from mice and atomic-bomb survivors is presented to understand the increasing patterns of cancer risks.

RESULTS

In both species, it was found that cancer mortality starts to increase earlier in the exposed group.

CONCLUSION

The results are consistent with the notion that in many irradiated organs (but not all) radiation-induced tissue damage can lead to the development of an altered microenvironment (most probably inflammation), which is favorable to the growth of spontaneously arising tumor cells and can lead to an earlier onset of the diseases or to an apparently increased risk of cancer.

ON RADIATION-INDUCED AGING: ACCELERATED OR PREMATURE AGING

The concept of radiation-induced aging is revisited from the viewpoint of a mathematical model. The effect of radiation on carcinogenesis is treated based on the Armitage-Doll multi-stage theory. The formula obtained for cancer incidence rate indicates that radiation dose can be explained in terms of time. Radiation-induced aging for acute and chronic exposures is described using age-specific cancer incidence rates as a measure of aging. It shows that accelerated aging is related to the dose rate, whereas premature aging is related to the cumulative dose, providing a simple and natural interpretation of radiation-induced aging. The usefulness of this approach is demonstrated by applying the formula to cancer prevalence data from mice chronically exposed to low dose-rate radiation.

MECHANISMS OF RADIATION CARCINOGENESIS: WHAT IS REALLY INDUCED?

It has been difficult to understand why the relative risk for cancer decreases with an increase in time since an exposure to radiation. It was recently recognized that this decline can be explained by a parallel shift of the age-related cancer mortality curve toward younger ages. In fact, it has been known for many years that mouse survival curves exhibit a parallel shift toward younger ages following an exposure to radiation, but it was not recognized that the mutation induction theory is incompatible with this parallel shift. This is because a parallel shift in the survival curve implies that all the irradiated individuals are affected, but the mutation induction theory assumes that only a fraction of the irradiated individuals is affected following an exposure to radiation. Thus, it seems likely that the target of radiation action, which leads to carcinogenesis, is not restricted to epithelial cells but includes all of the surrounding cells leading to an altered microenvironment. Since it is repeatedly observed that radiation-exposed normal tissues can stimulate transplanted or spontaneously arising tumor cells to grow faster, worsen the malignant phenotypes and finally kill the host earlier than usual, an exposure to radiation seems most likely to cause tissue inflammation, which creates conditions favorable for the growth of spontaneously arising tumor cells. This new concept suggests that it might be possible to attenuate the extent of radiation carcinogenesis by intervening in tissue inflammatory processes.

Differential expression of DNA-dependent protein kinase catalytic subunit in the brain of neonatal mice and young adult mice

It is generally thought that younger people are more susceptible to cancer development after exposure to ionizing radiation in reference to epidemiological studies and animal experiments. However, little is known about the age-dependent alteration in DNA repair ability. In the present study, we examined the expression levels of proteins involved in the repair of DNA double-strand breaks through non-homologous end joining (NHEJ), i.e., DNA-dependent protein kinase catalytic subunit (DNA-PKcs), X-ray repair cross-complementing 4 (XRCC4) and XRCC4-like factor (XLF). We found that the expression of DNA-PKcs in brain tissues was higher in neonatal mice (1 week after birth) than in young adult mice (7 weeks after birth). In association with this, DNA double-strand breaks were repaired more rapidly in the brain tissues of neonatal mice than in those of young adult mice. The current results suggested a possible role for DNA-PKcs protecting developing brain tissues from DNA double-strand breaks.

High susceptibility of mouse newborns to delayed appearance of DNA double-strand breaks in neural stem/progenitor cells exposed to ionizing radiation

Fetal brains are known to be extremely sensitive to ionizing radiation, which can induce structural and functional defects in the developing brain. However, there is less data on the effects of radiation on newborn brains. To determine the radiation sensitivity in newborn brains, we determined the number of DNA double-strand breaks (DSBs) appearing at later stage post-irradiation in neural stem/progenitor cells (NSPCs) of mouse newborns <3 days old, and compared it with the numbers of DSBs of fetal, 1-week-neonate, 2-week-neonate, and adult mice. DSBs in the nucleus were quantified by counting the number of foci of phosphorylated histone H2AX (γ-H2AX) in NPSCs using a newly developed computer program. Then, we irradiated 14-day fetuses, newborns <3 days old, 1-week-old neonates, 2-week-old neonates, and 12-week-old adult mice with 2 Gy of X-rays. At 6-7 weeks post-irradiation, the brain tissues isolated from the mice were incubated, and DSBs in the growing neurospheres were counted using a focus-counting program. The delayed appearance of DSBs by X-irradiation was evident in NSPCs derived from newborns <3 days old, as well as in 1-week-old neonates, 2-week-old neonates and adult mice, but not 14-day fetuses, at 6-7 weeks post-irradiation. It was of particular interest that the NSPCs of newborns were 2.5-fold more susceptible than those of adults to radiation-induced delayed appearance of DSBs, indicating that newborns <3 days old are the most vulnerable to the delayed effects of radiation among the mouse groups examined.

Age-dependent changes in DNA repair in radiation-exposed mice

Ionizing radiation (IR) is a well-known human carcinogen. Young and adult individuals are known to respond to radiation in a different manner. In this study, we analyzed changes in the spleen of juvenile (two-week-old), adult (two-month-old) and old (18-month-old) C57BL/6 male mice subjected to a whole-body exposure to 1 Gy of X rays. We measured the number of γ-H2AX foci and ATM protein levels as a reflection of the level of DNA double-strand breaks (DSBs), and found that old animals had a high frequency of occurrence of noninduced DSBs. Exposure to X rays resulted in a rapid increase in the number of DSBs in juvenile and adult animals at 6 h postirradiation followed by a return to preirradiated DSB values at 96 h postirradiation. No changes were observed in old animals. The analysis of the levels of proteins involved in DNA damage base excision and mismatch repair pathways, including KU70, RAD51, POL β, POL δ, POL ε, APE1 and MSH2 showed substantial age-dependent radiation-induced differences. Finally, we demonstrated that old animals had a higher background level of cell apoptosis compared to younger animals, but in contrast to younger animals, old animals were not able to commit spleen cells to apoptosis after being irradiated. Thus, spleen cells of old mice have a high level of spontaneous DNA damage, but they are not able to deal with additional radiation-induced damage as efficiently as younger animals, substantiating age-depending differences in radiation-induced DNA damage and repair response and its outcomes.

Direct measurement of a patient's entrance skin dose during pediatric cardiac catheterization

Children with complex congenital heart diseases often require repeated cardiac catheterization; however, children are more radiosensitive than adults. Therefore, radiation-induced carcinogenesis is an important consideration for children who undergo those procedures. We measured entrance skin doses (ESDs) using radio-photoluminescence dosimeter (RPLD) chips during cardiac catheterization for 15 pediatric patients (median age, 1.92 years; males, n = 9; females, n = 6) with cardiac diseases. Four RPLD chips were placed on the patient's posterior and right side of the chest. Correlations between maximum ESD and dose-area products (DAP), total number of frames, total fluoroscopic time, number of cine runs, cumulative dose at the interventional reference point (IRP), body weight, chest thickness, and height were analyzed. The maximum ESD was 80 ± 59 (mean ± standard deviation) mGy. Maximum ESD closely correlated with both DAP (r = 0.78) and cumulative dose at the IRP (r = 0.82). Maximum ESD for coiling and ballooning tended to be higher than that for ablation, balloon atrial septostomy, and diagnostic procedures. In conclusion, we directly measured ESD using RPLD chips and found that maximum ESD could be estimated in real-time using angiographic parameters, such as DAP and cumulative dose at the IRP. Children requiring repeated catheterizations would be exposed to high radiation levels throughout their lives, although treatment influences radiation dose. Therefore, the radiation dose associated with individual cardiac catheterizations should be analyzed, and the effects of radiation throughout the lives of such patients should be followed.

Lifetime increased cancer risk in mice following exposure to clinical proton beam-generated neutrons

PURPOSE

To evaluate the life span and risk of cancer following whole-body exposure of mice to neutrons generated by a passively scattered clinical spread-out Bragg peak (SOBP) proton beam.

METHODS AND MATERIALS

Three hundred young adult female FVB/N mice, 152 test and 148 control, were entered into the experiment. Mice were placed in an annular cassette around a cylindrical phantom, which was positioned lateral to the mid-SOBP of a 165-MeV, clinical proton beam. The average distance from the edge of the mid-SOBP to the conscious active mice was 21.5 cm. The phantom was irradiated with once-daily fractions of 25 Gy, 4 days per week, for 6 weeks. The age at death and cause of death (ie, cancer and type vs noncancer causes) were assessed over the life span of the mice.

RESULTS

Exposure of mice to a dose of 600 Gy of proton beam-generated neutrons, reduced the median life span of the mice by 4.2% (Kaplan-Meier cumulative survival, P=.053). The relative risk of death from cancer in neutron exposed versus control mice was 1.40 for cancer of all types (P=.0006) and 1.22 for solid cancers (P=.09). For a typical 60 Gy dose of clinical protons, the observed 22% increased risk of solid cancer would be expected to decrease by a factor of 10.

CONCLUSIONS

Exposure of mice to neutrons generated by a proton dose that exceeds a typical course of radiation therapy by a factor of 10, resulted in a statistically significant increase in the background incidence of leukemia and a marginally significant increase in solid cancer. The results indicate that the risk of out-of-field second solid cancers from SOBP proton-generated neutrons and typical treatment schedules, is 6 to 10 times less than is suggested by current neutron risk estimates.

Induction and persistence of radiation-induced DNA damage is more pronounced in young animals than in old animals

Younger individuals are more prone to develop cancer upon ionizing radiation (IR) exposure. Radiation-induced tumors are associated with inefficient repair of IR-induced DNA damage and genome instability. Phosphorylation of histone H2AX (γ-H2AX) is the initial event in repair of IR-induced DNA damage on the chromatin flanking the DNA strand breaks. This step is crucially important for the repair of DNA strand breaks and for the maintenance of genome stability. We studied the molecular underpinnings of the age-related IR effects using an animal model. By assaying for IR-induced γ-H2AX foci we analyzed the induction and repair of the DNA strand breaks in spleen, thymus, liver, lung, kidney, cerebellum, hippocampus, frontal cortex and olfactory bulb of 7, 14, 24, 30 and 45 days old male and female mice as a function of age. We demonstrate that tissues of younger animals are much more susceptible to IR-induced DNA damage. Younger animals exhibited higher levels of γ-H2AX formation which partially correlated with cellular proliferation and expression of DNA repair proteins. Induction and persistence of γ-H2AX foci was the highest in lymphoid organs (thymus and spleen) of 7 and 14 day old mice. The lowest focal induction was seen in lung and brain of young animals. The mechanisms of cell and tissue-specificity of in vivo IR responses need to be further dissected. This study provides a roadmap for the future analyses of DNA damage and repair induction in young individuals.

The balance between initiation and promotion in radiation-induced murine carcinogenesis

Studies of radiation carcinogenesis in animals allow detailed investigation of how the risk depends on age at exposure and time since exposure and of the mechanisms that determine this risk, e.g., induction of new pre-malignant cells (initiation) and enhanced proliferation of already existing pre-malignant cells (promotion). To assist the interpretation of these patterns, we apply a newly developed biologically based mathematical model to data on several types of solid tumors induced by acute whole-body radiation in mice. The model includes both initiation and promotion and analyzes pre-malignant cell dynamics on two different time scales: comparatively short-term during irradiation and long-term during the entire life span. Our results suggest general mechanistic similarities between radiation carcinogenesis in mice and in human atomic bomb survivors. The excess relative risk (ERR) in mice decreases with age at exposure up to an exposure age of 1 year, which corresponds to mid-adulthood in humans; the pattern for older ages at exposure, for which there is some evidence of increasing ERRs in atomic bomb survivors, cannot be evaluated using the data set analyzed here. Also similar to findings in humans, initiation dominates the ERR at young ages in mice, when there are few background pre-malignant cells, and promotion becomes important at older ages.

A preliminary operational classification system for nonmutagenic modes of action for carcinogenesis

This article proposes a system of categories for nonmutagenic modes of action for carcinogenesis. The classification is of modes of action rather than individual carcinogens, because the same compound can affect carcinogenesis in more than one way. Basically, we categorize modes of action as: (1) co-initiation (facilitating the original mutagenic changes in stem and progenitor cells that start the cancer process) (e.g. induction of activating enzymes for other carcinogens); (2) promotion (enhancing the relative growth vs differentiation/death of initiated clones (e.g. inhibition of growth-suppressing cell-cell communication); (3) progression (enhancing the growth, malignancy, or spread of already developed tumors) (e.g. suppression of immune surveillance, hormonally mediated growth stimulation for tumors with appropriate receptors by estrogens); and (4) multiphase (e.g., "epigenetic" silencing of tumor suppressor genes). A priori, agents that act at relatively early stages in the process are expected to manifest greater relative susceptibility in early life, whereas agents that act via later stage modes will tend to show greater susceptibility for exposures later in life.

Dose-response relationship for induction of ovarian tumors in mice irradiated during prenatal, early postnatal and elder periods

Female B6C3F1 mice were irradiated at day 17 prenatal period, day 0, 7, 35, 105, 240, 365 and 550 postnatal period with doses of 0.10 to 5.70 Gy gamma rays from (137)Cs. All mice were allowed to live through their entire life spans under a specific-pathogen free condition. The excess relative risk for prevalence at the time of death of ovarian tumors was used as a comprehensive measure of radiation effect. The excess relative risks at 1 Gy were estimated for all irradiated groups based on the dose-response relationships and compared to each other. A marked increase in susceptibility was found during the age between day 17 prenatal and day 0 postnatal period. A drastic decrease in susceptibility was observed during the period between day 105 and day 240. The shape of the dose-response curve was downward concave in mice irradiated at day 0, 7, 35 or 105 postnatal period, whereas, the downward curvature of dose-response was not observed in mice irradiated at day 17 prenatal period, day 240, 365 or 550 postnatal period. It has become obvious that mice of the early postnatal, pre-puberty and young adult periods are highly susceptible to induction of ovarian tumors by gamma rays.

X radiation up-regulates the occurrence and the multiplicity of invasive carcinomas in the intestinal tract of Apc(min/+) mice

X rays are well known to cause genetic damage and to induce many types of carcinomas in humans. The Apc(min/+) mouse, an animal model for human familial adenomatous polyposis (FAP), contains a truncating mutation in the APC gene and spontaneously develops intestinal adenomas. To elucidate the role of X rays in the development of intestinal tumors, we examined the promotion of carcinogenesis in X-irradiated Apc(min/+) mice. Forty out of 77 (52%) X-irradiated Apc(min/+) mice developed adenocarcinomas that invaded the proprial muscle layer of the small intestine; 24 of 44 (55%) were in males, and 16 of 33 (49%) were in females. In contrast, invasive carcinomas were detected in the small intestines of only 13 of 64 (20%) nonirradiated Apc(min/+) mice; nine of 32 (28%) were in males and four of 32 (13%) were in females. These differences between X-irradiated and nonirradiated Apc(min/+) mice in the occurrence of invasive intestinal carcinomas were statistically significant (P < 0.05 for males, P < 0.005 for females). In wild-type mice, invasive carcinomas were not detected in either X-irradiated or nonirradiated mice. Apc(min/+) mice had many polyps in the large intestine with or without X irradiation; there was no difference in the number of polyps between the two groups. Also, invasive carcinomas were not detected in the large intestine with or without irradiation. The occurrence of mammary tumors, which was observed in Apc(min/+) mice, was found to be increased in irradiated Apc(min/+) mice (P < 0.01). Apc(min/+) mice had many polyps in the small and large intestines with or without X irradiation. X-irradiated Apc(min/+) mice had highly invasive carcinomas in the small intestine with multiplicities associated with invasiveness. Our results suggest that X radiation may promote the invasive activity of intestinal tumors in Apc(min/+) mice.

Cause of Death and Neoplasia in Mice Continuously Exposed to Very Low Dose Rates of Gamma Rays

Four thousand 8-week-old SPF B6C3F1 mice (2000 of each sex) were divided into four groups, one nonirradiated (control) and three irradiated. The irradiated groups were exposed to (137)Cs gamma rays at dose rates of 21, 1.1 and 0.05 mGy day(-1) for approximately 400 days with total doses equivalent to 8000, 400 and 20 mGy, respectively. All mice were kept until natural death, and pathological examination was performed to determine the cause of death. Neoplasms accounted for >86.7% of all deaths. Compared to the nonirradiated controls, the frequency of myeloid leukemia in males, soft tissue neoplasms and malignant granulosa cell tumors in females, and hemangiosarcoma in both sexes exposed to 21 mGy day(-1) were significantly increased. The number of multiple primary neoplasms per mouse was significantly increased in mice irradiated at 21 mGy day(-1). Significant increases in body weights were observed from 32 to 60 weeks of age in males and females exposed to 1.1 mGy day(-1) and 21 mGy day(-1), respectively. Our results suggest that life shortening (Tanaka et al., Radiat. Res. 160, 376-379, 2003) in mice continuously exposed to low-dose-rate gamma rays is due to early death from a variety of neoplasms and not from increased incidence of specific neoplasms.

In utero and neonatal sensitivity of ApcMin/+ mice to radiation-induced intestinal neoplasia

PURPOSE

To assess the sensitivity of ApcMin/+ mice (adenomatous polyposis coli Apc, multiple intestinal neoplasia, Min) to the development of intestinal adenomas after x-irradiation in utero, as neonates, or as young adults.

MATERIALS AND METHODS

CHB6 ApcMin/+ mice were exposed to an acute dose of 2 Gy x-rays either in utero on day 7 or 14 post-conception, as 2-day or 10-day neonates or as 35-day young adults. Tumour identification and counting was performed 200-214 days later.

RESULTS

Irradiation as 10-day-old neonates resulted in a significantly greater overall tumour incidence (average of about 130 tumours per animal) than irradiation as 35-day-old young adults (about 70 tumours). Irradiation as 2-day-old neonates resulted in an intermediate incidence (about 85 tumours). In contrast, the greatest tumour incidence observed after in utero irradiation of ApcMin/+ mice, of about 44 tumours per animal after 2 Gy irradiation at 14 days post-conception, was significantly lower than the incidence in irradiated adults. Tumour incidences after irradiation as 7-day embryos was not significantly raised above numbers in unirradiated controls (about 30 tumours). These tumour numbers include cystic crypts, largely radiation-induced, which were classed as early stage microadenomas on the basis of loss of wild-type Apc+ and expression of beta-catenin.

CONCLUSIONS

The sensitivity of ApcMin/+ mice to the induction of intestinal tumours by radiation was shown to be in the order: 10 d neonates>2 d neonates>35 d young adults>14 d fetus>7 d embryo.

Dose-response relationship for life-shortening and carcinogenesis in mice irradiated at day 7 postnatal age with dose range below 1 Gy of gamma rays

This study was designed to elucidate the dose-response relationships for life-shortening and tumorigenic effect in the dose range below 1 Gy of gamma rays delivered during the infant period. Female B6C3F1 mice were irradiated with 0.10, 0.48 or 0.95 Gy at 7 days of age. All irradiated mice were allowed to live out their entire life span together with a simultaneously ongoing control group under a specific pathogen-free condition. Shortening of the mean life span was 1.58% in mice irradiated with 0.10 Gy, which was statistically significant . The coefficient of the linear dose-response relationship for life-shortening was 11.21% Gy(-1). The attributable death fraction for all causes of death in 0.10 Gy group reached 0.092. The excess relative risk for death rate from all causes was 0.102 in the group irradiated with 0.10 Gy. The coefficient of the linear dose-response relationship of the excess relative risk for death rate from all causes was 1.30 Gy(-1). The mean number of types of solid tumors at the time of death in mice irradiated with 0.10 Gy was distinctly larger than that in the control group. The excess relative risk for death rate from solid tumors was 0.45 in mice irradiated with 0.10 Gy. The coefficient of the linear dose-response relationship of excess relative risk for death rate from solid tumors was 4.52 Gy(-1). Increase in incidences of the pituitary, ovarian and adrenal tumors was observed in mice irradiated with 0.10 Gy. The results of the present study showed that infant mice are susceptible to solid tumor induction, especially of the endocrine organs.