In toxicology, the lethal dose (also referred to as LD₅₀, lethal dose 50% or median lethal dose) is an indicator used as a measure of cell killing and fertility. The LD₅₀ for a particular substance (i.e., ionizing radiation) refers to the amount (mgr) necessary per weight (kg) to cause death in 50% of the population under study. Early studies used this measure of toxicological effects of ionizing irradiation in mouse [16,58], rat [16,19,20,64], Chinese hamster [33,34,35] and guinea pig [27,28,29,30]. Results have been heterogeneous depending on the species studied and the moment of the irradiation (oocytes and pachynema, diplonema-dictionema or at the moment of birth). Although studies are scarce, the observed trend is that LD is species-dependent.

Fertility alterations such as oocytes depletion, ovarian failure and loss of reproductive potential have also been used as indicators of radiation-induced genotoxic effects only in a few numbers of reports (Table 1). This includes studies in mouse [16,22], guinea pig [27,28,29,30] and rat [39,40,41]. The results so far are contradictory but the general observed trend is that fertility alterations do not seem to be related to the species irradiated but rather to the irradiation timing.

Developmental malformations are anatomical or structural abnormalities observed at birth and may be caused by genetic factors or environmental alterations (or a combination of the two) that occur during prenatal development. Although genetic factors are the most common causes of developmental malformations, the underlying mechanisms are complex and not well known. Due to this complexity, only few authors have used developmental malformations as indicators of radio-induced insults (Table 1). At the same time, the results obtained so far (either positive or negative) can even be contradictory due to the great diversity of alterations that can be considered as “malformations”, as well as to different moments (pre- and post-implantation, for example) when observed.

The DNA double-strand breaks (DSBs) produced by IR constitute a particular hazard to cells and can result in genetic mutations that can be evaluated. Several parameters are used to evaluate radiation-induced genetic mutations, and here we will review those that can directly affect the DNA (i.e., dominant lethal effects, locus-specific mutations, chromosomal aberrations, genomic instability) or associated proteins (i.e., synaptonemal complex).

A parameter that has been traditionally used in toxicology has been the dominant lethal effects, (DL) [44,45,57]. DL events are genetic effects cause by any physical or chemical agent that causes embryonic or fetal death, indicating that the substance has affected germinal tissue. According to the OECD (Organization for Economic Co-operation and Development) guidelines for the testing of chemicals, “a dominant lethal mutation is one occurring in a germ cell which does not cause dysfunction of the gamete but which is lethal to the fertilized egg or developing embryo”. Therefore, DL causes embryonic or fetal death. They are generally accepted to be the result of chromosomal damage, but gene mutations and toxic effects cannot be excluded. Studies performed in golden hamster, mouse and guinea pig [44,45,57] have revealed that the radiation sensitivity obtained from evaluating DL is both species- and oocytes stage-dependent (Table 1).

Additionally, locus-specific mutations have been used to detect recessive mutations induced in diploid organisms: a strain that carries several known recessive mutations in a homozygous condition is crossed with a non-mutant treated strain in order to look for these recessive mutations in the progeny. This has been the case of mutations in genes controlling mouse fur color, a method initially developed by Russell and Russell [65] and applied later on. However, this method has been reported to have low mutation rates and therefore low sensitivity to detect trans-generational effects [66]. An alternative to overcome this limitation is the detection of loci containing repetitive sequence elements (expanded simple tandem repeat—ESTR—DNA loci), which are known to mutate at an extremely high frequency [67]. Jointly with chromosomal reorganizations, this has been the technique most widely used to detect trans-generational genotoxic effects of chemical and physical (i.e., IR) agents, as we shall see later on.

When IR damages DNA and induces chromosomal anomalies through the formation of DSBs, such alterations can be stabilized through chromosomal rearrangements or capped by the addition of telomeric sequences, producing chromosomal aberrations. Revisions focused on different types of radiation-induced chromosomal reorganizations previously have been reported elsewhere [68], so here a brief summary is provided. As a general overview, chromosomal reorganizations induced by IR are mainly structural changes resulting in inversions (either paracentric or pericentric), translocations, fusions and fissions, among others, and to a lesser extent, numerical. These structural aberrations can be considered as (i) chromatid-type or (ii) chromosome-type. The former affects one of the sister-chromatids, whereas both sister-chromatids are involved in the latter. Both types of chromosomal aberrations have been used as reliable biomarkers for genetic radio-induced damage. When occurring in the germ line, chromosomal aberrations can lead to aneuploidy and/or constitutional aberrations in the offspring.

In a genome-wide broader scale, the increased rate of alterations produced by the effect of IR is known as radiation-induced genomic instability (RIGI) [69]. Genomic instability is a term initially used to describe a phenomenon that results in the accumulation of multiple changes required to convert a stable genome of a normal cell into an unstable genome characteristic of a tumor, and it is known that this can be caused by radiation [70]. RIGI can be measured by the formation of gene amplification, micronucleus formation and microsatellite instability, as well as chromosomal reorganizations, and its effects are dependent on the genetic background of the cell [71]. When inducing genetic abnormalities that can be propagated in the offspring, “trans-generational RIGI” can also be measurable as a biomarker for radio-induced genotoxic effects [41].

Radiation-induced genetic mutations can also be measured by analyzing the proteins associated with the DNA. In the case of the germ line, the synaptonemal complex (SC) has been used as a biomarker for genotoxic effects [20,21,40]. SC is a meiosis-specific protein structure that promotes synapsis (pairing of homologous chromosomes) and recombination events during the first meiotic prophase [72,73,74]. SC consists of two parallel lateral elements (LE), one central element (CE), and numerous transverse filaments (TF) (reviewed in [74]). The Synaptonemal Complex Proteins 2 and 3 (SYCP2 and SYCP3) are layered onto cohesin proteins producing the lateral elements of SC during the first meiotic prophase, whereas the transverse filament protein SCP1 [75] bridges the lateral elements at pachynema. Recent results have shown that the synaptonemal complex is required for both synapsis and the structural integrity of the chromosome axis [76]. Moreover, the repair of meiotic DSBs is impaired in the absence of SC in mice, highlighting the importance of this structure for the correct progression of meiosis in mammals [76]. Cytogenetic studies have used the synaptonemal complex as a biomarker for the genotoxic effects induced by IR, given that any chromosomal aberration can be detected when the SC is completely assembled during the pachytene stage [20,21,40,77]. Early studies performed by Allen and collaborators [78] detected three main types of SC abnormalities induced by chemical mutagens in male mice: (i) synaptonemal complex fragmentation, (ii) synaptic abnormalities and (iii) mispairing, or heterosynapsis. Later studies have identified other types of abnormalities, in most cases stage-dependent, of which the most common ones are chromosomal reorganizations that can be observed as circles, single or double, multivalents and loops [20,77,78]. Moreover, alterations in synapsis can also be also detectable, such as partial and/or complete asynapsis (reflected by the presence of univalents or bivalents with unpaired regions) or non-homologous pairing [19]. Very few studies have been focused on the radio-induced genotoxic effects in early stage of meiosis (oogonia and prophase I oocytes) and they are restricted to rat [19,20,39,40], mouse [21], Chinese hamster [33] and guinea pig [27]. Out of these, the studies performed in rat and mouse [19,20,21,39,40] have used SC as an indicator for genotoxic effects.

Most of our knowledge regarding the radiation sensitivity of the female germ cells comes mainly from experiments performed in mice and, to a lesser extent, in rat, guinea pig and Chinese hamster. Radiation-induced genotoxic effects have been studied throughout all meiotic stages: oogonia, oocyte I (at leptonema, zygonema, pachynema or diplonema), immature oocytes (oocytes I at primordial or unilaminar follicles), maturing oocytes (oocytes I at pre- and antral follicles) and mature oocytes (oocytes I/oocytes II at pre- and ovulatory follicles) (Figure 1, Table 1). Moreover, there are also studies focused on post-conception stages, mainly at the zygote and pre-implantation stages (1-cell to blastocyst) and post-implantation stages (Table 1). All of these studies indicate that oocyte radiation-sensitivity depends on its developmental stage and its relation with folliculogenesis.

As already stated, few studies have focused on oogonia and prophase I oocytes [19,20,21,39,40]. Pujol and collaborators [19,20] analyzed the effect of different doses of X-rays (1 Gy, 2 Gy and 5 Gy) on the number of recoverable germ cells in female rat fetuses irradiated at three different days of gestation (d.g.): 14 d.g. (oogonia), 16 d.g. (leptonema) and 18 d.g. (zygonema). At 14 d.g., all germ cells are at the stage of mitotic proliferation, and as a result many oogonia are expected to be at the S phase of mitosis. At 16 d.g., many cells are completing their last mitotic division, and others have already entered the first meiotic prophase, whereas at 18 d.g. most cells have reached leptonema or lepto-zygonema. Their results indicate that irradiation significantly decreases the number of germ cells, and increases the incidence of synaptonemal complex fragmentation, suggesting that differences in sensitivity to irradiation depend on the meiotic stage at the time of irradiation. Differences in radiation doses were also observed; the 5 Gy dose is lethal for oogonia and zygotene oocytes, while for leptotene oocytes, this dose could have a sub lethal effect. Overall, Pujol and collaborators [19,20] show that SC fragmentation is stage-dependent; they found a significant increase in the incidence of heterosynapses in oogonia, a significant increase in the frequency of structural anomalies in leptotene oocytes and a significant increase in nuclear fragmentation in zygotene oocytes. Similar results were obtained by Johannisson and co-workers [21], when fetal female mice were exposed to ionizing irradiation of 2 Gy in a single dose at 14, 16, and 17 d.g., indicating that at early stages of meiosis, the genotoxic effects of IR are not species-dependent.

Studies on immature and mature oocytes are more abundant in the literature. Overall, the observed general view is that mature oocytes are more radio-sensitive than are immature oocytes. These observations were already reported by early studies in mouse [16,37,46,54,55,65,80]. Later on, this trend was confirmed in other model species such as rat and guinea pig. Studies conducted in rat [39,40] showed that pre-pubertal dictiotene oocytes are less radio-sensitive than are post-pubertal dictiotene oocytes. In the same vein, and according to studies performed by Jacquet’s lab in guinea pig [23,30,79], immature oocyte are less radiation-sensitive than are maturing oocytes, and these are less radiation-sensitivity than are mature oocytes. On the other hand, it has been reported that the chromosomal sensitivity of immature dictiotene oocytes enclosed in primordial follicles is very low in the guinea pig [45]. In the late pre-antral stages the sensitivity of the oocyte gradually increases to reach a maximum at the antral stage of folliculogenesis. The same occurs in the Chinese hamster, where oocytes surrounding ovulation show the greatest sensitivity to radiation [62].

During oocyte maturation (between one day and six weeks after irradiation in the mouse), regardless of the fact that in all cases oocytes are arrested at the dictiotene prophase I stage, differences in radiation sensitivity have been reported. In mouse, ovulated oocytes (MII stage) analyzed one day to two weeks post-irradiation are less sensitive to the induction of chromosomal aberrations, genetic mutations and dominant lethal effects (DL) than are ovulated oocytes analyzed after three weeks and six weeks post-irradiation [47,57,80]. In particular, Tease [47] observed that at medium/high doses (100 cGy–600 cGy) radio-induced chromosomal aberrations increased two-fold in immature oocytes in relation to mature oocytes (Table 1). Ovulation is another key point. A few hours before ovulation, pre-ovulatory oocytes are more radio-sensitive, given that some studies have reported a significant increase in chromosomal aberrations [37,48,49,56] and chromosomal non-disjunction [81] in mouse irradiated at low doses (0.1 Gy–0.5 Gy). Analyzing DL, other authors have found that pre-ovulatory oocytes are radiosensitive in mouse [65,82] and rat [83]. Therefore, the sensitivity to the induction of genetic damage is dependent on the stage of folliculogenesis and, in this sense, the interval between irradiation and ovulation.

At this point and taking into account published data on mammalian females (human included), we can consider that the genetic radiation-sensitivity observed in the human female is related to the size of the follicles in the way that small follicles are apparently less sensitive to cell-killing than are large follicles. In the case of mouse germ cells, radiation-sensitivity is related to the folliculogenesis stage (Figure 2) with its maximum sensitivity in early diakinesi (as seems to happen in Chinese hamster) and its minimum in the full antral stage. The few data published on the human female [11,12,16] differ from those obtained in guinea pig, where the maximum sensitivity would be in the full antral stage (Figure 2). Pre- and antral follicles have a larger (both in number and volume) granulose layer than do primordial follicles, probably acting as a protective screen against the genotoxic effects of radiation. Therefore, why, in some species, are mature oocytes more radio-sensitive than are immature oocytes? One has to take into account that a follicle works as a functional syncytium, so the response may lie in how the DNA damage-response changes during oocyte development and how they are challenged by ionizing irradiation. Moreover, we have to bear in mind, however, that all parameters used in the literature to measure radiation-induced genotoxic effects are highly diverse (from chromosomal alterations to cell-killing and congenital malformations) and, therefore, the repair mechanisms involved might be different so, therefore, we have to be extremely cautious when extrapolating the data. Notwithstanding these differences, a wide variety of direct and indirect evidence indicates that oocytes maintain their capacity to repair DNA insults during their maturation process [84,85,86,87,88,89,90]. In a recent study, Wang and collaborators [91] described the proteome of mouse oocytes at different developmental stages, the dictiotene or germinal vesicle (GV) stage, the metaphase II (MII) stage and zygote stage (fertilized oocyte). Surprisingly, the authors found that more DNA repair proteins are expressed in MII oocytes than in GV oocytes or the zygote; they included in the analysis the proteins involved in the SSB repair and nucleotide-excision repair pathways. Moreover, they found that the DNA recombination pathway and DNA replication protein families are also more abundant in MII oocytes. In particular, 35 out of 53 proteins involved in the DNA repair process identified in both MII oocytes and zygotes are up-regulated in the MII oocytes. These include Rad51, BRCA2, XRCC1, PARP1 and TRIP13. These results are consistent with previous findings on gene expression profiles of oocytes and early embryos [92], which suggest that the over-representation of genes involved in DNA repair confer protection to maintain genome stability of the female germ-line.

Based on the evidence, a possible interpretation of differences in radiation-sensitivity during oocyte maturation is that mature oocytes repair genetic insults by means of “prone to error” mechanisms, stabilizing alterations through chromosomal rearrangements. That would explain why chromosomal aberrations are frequently observed at MII [37,49,50,93]. On the contrary, immature oocytes would not efficiently repair the genetic insults, resulting in cell death and not contributing to the offspring. Species differences to the genotoxic effects of IR can also be attributed to differences in the implementation of DDR and, therefore, more studies are needed to overcome any disparity in reports.

When an increased rate of mutation or augmentation of developmental malformations, to name two different measurable endpoints, is observed in the non-exposed offspring of irradiated parents, we are facing trans-generational genomic instability [37,49,50,93]. According to Barber and Dubrova [94], the first evidence of a trans-generational effect for ionizing radiation was demonstrated by Luning et al. [95], who observed elevated rates of dominant lethal mutations in the non-exposed first-generation (F₁) offspring of males injected with 239^Pu citrate solution. Data on potential trans-generational genomic instability following irradiation are largely limited to males [94,96,97,98,99], and only until very recently did studies in females begin to flourish [39,40,41,42,43,51,60]. Early studies evaluated the genotoxic effects in prophase I stages (pachynema, diplonema and dictionema oocytes) in the Chinese hamster [33,34,35]. After exposing animals to a single dose of 1 Gy, the authors did not observe a significant increase of chromosomal aberrations in one-cell embryos and/or developmental alterations among the progeny of the irradiated females. Therefore, and although these studies detected that radio-resistance was stage-dependent, they did not observe any trans-generational effects with these measurable endpoints. Subsequent studies in mice by Pils and collaborators [22] studied the genotoxic effects when irradiating females at the zygote stage. They not only obtained positive results when looking for fertility alterations of these irradiated females, but they also found an increment of congenital malformations in the F₁ progeny. Nevertheless, years later, Jaquet and co-workers [24] reported negative findings regarding developmental defects and genomic instability (analyzing chromosomal aberrations) in the F₁ and after exposure of low doses of irradiation at the mouse female zygote stage (0.2 and 0.4 Gy instead of 1 Gy used by Pils et al. [22]). Here, the authors suggested differences in the mouse strain and sex-radiation sensitivities. Interestingly, studies leaded by Dubrova’s group [51,60], based on the analysis of expanded simple-tandem repeats (ESTR), indicate that whereas paternal irradiation destabilizes the genomes of non-exposed offspring, maternal irradiation does not affect stability of their offspring. Taking together, these data show that the trans-generational effects in the offspring of irradiated male mice may be explained by a genome-wide destabilization, which manifests in many, possibly all, tissues. Why this behavior regarding ESTR is not observed in females needs further validation.

Others have studied trans-generational genomic instability through the analysis of chromosomal alterations detected in fibroblasts derived from fetuses, which resulted from fertilization of irradiated oocytes [25,26,39,40,41,42,43]. In their initial studies using mouse as a model species, Pampfer and Streffer [25] assumed that the observed structural chromosomal aberrations were not induced directly by radiation exposure and the radiation exposure of the zygote has led to instability of the genome so that “spontaneous” aberrations occurred more frequently. In doing so, these authors had already detected an increase in trans-generational chromosomal aberrations in mouse zygotes and, although these former results were not totally conclusive, “the intuited” trans-generational chromosomal instability in mouse was corroborated by Streffer, [26] some years later. To our knowledge, there is only one published work in mouse with trans-generational positive results [49]. In this case the statically significant F₁ chromosomal aberrations only correspond to pre-implantation stages as if the chromosomal aberrations were life-incompatible with later stages.

Mirroring the commented effects in mouse, studies in rat performed by our group came to similar conclusions. The rat has been used as a model species for evaluating the radiation-induced genotoxic effects in mammalian female germ cells, either in the irradiated females or in the offspring [19,20,39,40,41,42,43]. In general terms, studies of trans-generational radiation-induced genotoxic effects on the gonadal function of rat females, using mainly SCs or chromosomal alterations as measurable endpoint markers, have concluded that, in the case of F₀ fetal oocytes, cell-killing is dose-dependent and that SC fragmentation is prophase-stage-dependent [19,20]. Moreover, no trans-generational cytogenetic effects were observed in the progeny of irradiated females, indicating that those oocytes that remain after irradiation do not present genome instability or any constitutional aberration in the SCs. The same occurs when studying the F₁ fetal oocytes from pre-pubertal and adult irradiated mothers by SCs where the remaining oocytes do not present either genome instability or any constitutional aberration, although there is an induction of apoptosis in the fetal rat germ cells [39,40]. More recently, we studied fibroblasts derived from fetuses of irradiated mothers [41,42,43], and in this case, a radiation-induced chromosomal instability was observed (an increase of chromosomal instability but not constitutional chromosomal alterations) in the offspring of irradiated females. In our view, this apparent contradiction to previous data [39,40] suggests that the use of chromosomal parameters and not just the analysis of SCs is more efficient in detecting germ cell genomic instability.