Despite all the information gathered in the almost 25 years since its discovery, it is still not known how the GRC is passed down through the generations. One reason for this is that even though there is a lot of information about the behavior of the GRC in spermatogenesis and in the early stages of female meiosis, knowledge about its behavior during early embryo development and late stages of female gametogenesis is still lacking. In an attempt to explain the differences in GRC copy number between sexes (as well as the occasional variation in GRC copy number in the same sex or even within a single individual), possible scenarios of GRC inheritance have been proposed that will be discussed below.
Maternal inheritance of a single GRC copy
Cytogenetic studies of the GRC have predominantly focused on examining the GRC during meiosis (Fig. 3a). In males, the GRC is normally observed as a single univalent chromosome in spermatogonia (Pigozzi and Solari 1998; del Priore and Pigozzi 2014) and primary spermatocytes (Pigozzi and Solari 2005; Torgasheva et al. 2019). This single GRC is eliminated from the nucleus during the first meiotic division and forms a micronucleus that is later ejected from the cell (Pigozzi and Solari 1998, 2005; del Priori and Pigozzi 2014). In females, the GRC is usually found as two copies (i.e., a paired bivalent, resembling autosomes) that engage in regular crossing-over (Fig. 1b; Pigozzi and Solari 2005; del Priori and Pigozzi 2014). These observations led to the view that the GRC is only inherited from the maternal side (Fig. 3a). A case of maternal inheritance of the GRC was recently reported in F1 hybrids between two different munia species (Lonchura spp.) with distinct GRC size, where the F1 hybrids’ GRC matched the size of the maternal species’ GRC (Sotelo-Muñoz et al. 2022). In addition, sequencing of germline samples (testis and ejaculates) from multiple male zebra finches from the same family showed that all brothers shared the same GRC haplotype as the brother of their mother (Pei et al. 2022). However, the apparent uniparental inheritance of a single GRC copy raises the question about when and how a second copy arises in the female, but not in the male.
Pigozzi and Solari (2005) suggested that the sex difference in GRC copy number arises during the germline/soma differentiation. They proposed that in females, nondisjunction of the GRC sister chromatids during a mitotic division may produce a germline progenitor cell with two GRC copies and a somatic progenitor cell with no GRC. In males, one GRC chromatid may be transmitted to the germline progenitor cell and the other one lost (presumably by chromatid lagging during mitotic anaphase), leading to a somatic progenitor cell without a GRC, and a germline with a single GRC (Fig. S1). Pigozzi and Solari (2005) hypothesized that sex-biased expression of cohesin-related genes might be involved in these differences in chromosome behavior. This explanation provides a mechanism to explain not only sex differences in GRC copy number but also how the GRC is lost from somatic cells.
However, it is also possible that the sex-specific difference in GRC copy number does not arise during the differentiation between germline and soma (Fig. 3b), but at a later stage, for example, during mitotic divisions of primordial germ cells (PGCs; Fig. 3c). Nondisjunction of GRC sister chromatids during the mitosis of female PGCs might be a simple mechanism that could lead to one cell without the GRC (which might undergo apoptosis or become a somatic cell) and another cell with two GRC copies that are able to synapse and recombine during meiosis. Such an asymmetrical cell division with GRC non-disjunction would need to be strictly regulated to occur only once and only in females.
Occasional paternal inheritance of the GRC
While maternal inheritance appears to be the norm, very recent studies suggest that the GRC can occasionally be paternally inherited. Pei et al. (2022) demonstrated that the GRC can occasionally be paternally inherited based on three main findings: (1) a hybrid individual between the two zebra finch subspecies exhibiting mitochondrial DNA from the maternal subspecies but a GRC from the paternal subspecies, (2) a striking topological incongruence between mtDNA and GRC haplotype trees, suggesting that at least some GRC haplotypes were able to cross matriline boundaries, and (3) the presence of the GRC in a small portion of sperm heads. Interestingly, Pei et al. (2022) found that males varied substantially in the proportion of spermatozoa (1–19%) that contained the GRC, and that this pattern is family-specific, with males from the same family showing a consistently low or high proportion of GRC retention in their spermatozoa. This suggests a heritable component for GRC presence in sperm cells. The GRC was also observed in spermatozoa of great tits, although it was rare (in 3 out of 880 spermatozoa; Torgasheva et al. 2021).
It is currently unclear what happens when the zygote receives two copies of the GRC, one of maternal and one of paternal origin. Pei et al. (2022) did not observe any GRC-heterozygous individuals in their sample, but this may simply be a consequence of biparental inheritance being uncommon. If receiving two copies of the GRC would cause substantial problems during embryonic development, such paternal inheritance should be selected against. However, if no problems arise, the strategy of biparental inheritance should rapidly outcompete the strategy of restriction to maternal inheritance, simply because GRCs that are biparentally inherited are more likely to be passed on to the next generation. The observed polymorphism and high repeatability in the effectiveness of GRC elimination during spermatogenesis open up the possibility to study and evaluate the success of the paternal inheritance strategy in the future.
Polymorphism and mosaicism in GRC copy number and size
The GRC is usually present as two copies in oocytes and as a single copy in spermatocytes (Pigozzi and Solari 2005). However, polymorphism (i.e., variation among individuals) and mosaicism (i.e., variation within the same individual) in GRC copy number have been observed in males and females of some species (Table 1). In the zebra finch (Pigozzi and Solari 1998, 2005) and the sand martin Riparia riparia (Fig. 1c; Malinovskaya et al. 2020a), female individuals with a single GRC copy in all their primary oocytes have been found, although in a relatively low proportion (12% of zebra finch females and 17% of sand martin females; Table 1). In the great tit Parus major, four of seven females showed mosaicism for GRC copy number, with the majority of primary oocytes containing two GRC copies and the minority (from 2 to 26%) a single copy (Torgasheva et al. 2021). These observations might be explained by the failure of GRC duplication in some or all PGCs during female embryonic development (Fig. 3c). It is plausible that GRC duplication might not be absolutely essential, assuming that in females with a single GRC copy, the unpaired GRC univalent might remain in the egg cell while the polar body does not receive any GRC. The non-negligible frequency in which this has been observed in zebra finch and sand martin populations might mean that females with a single GRC copy do not have a dramatically reduced fitness (compared to females with two GRC copies) and that the duplication of the GRC in females is not under strong selection pressure.
In males, mosaicism in GRC copy number has been observed in four species (Table 1). In pale martins, Riparia diluta, seven out of nine analyzed males showed GRC copy number mosaicism in primary spermatocytes (Fig. 1d; Malinovskaya et al. 2020a). In these males, most primary spermatocytes had a single GRC copy, but spermatocytes with two or even three copies were also observed. Similar observations were described in the great tit (Torgasheva et al. 2021) and black-headed munia (Sotelo-Muñoz et al. 2022), wherein some spermatocytes with two GRC copies were observed. In one of the two analyzed individuals of European pied flycatcher, most of the primary spermatocytes surprisingly carried two GRCs (Slobodchikova et al. 2022). The relatively high number of species in which such mosaicism has been observed could suggest that the segregation of the GRC during mitosis is often unstable in males. Occasional non-disjunction of GRC sister chromatids during mitotic divisions of PGCs might be the reason why some spermatocytes carry more than a single GRC copy, as it has been observed for B chromosomes in some species (Nur 1963; Jones 2018). It is plausible that such variability may be inconsequential since all GRC copies seem to be canonically eliminated during spermatogenesis (Pigozzi and Solari 2005; del Priori and Pigozzi 2014; Sotelo-Muñoz et al. 2022).
Polymorphism and mosaicism were observed not only for GRC copy number but also for GRC size. In one male black-headed munia, a small proportion of spermatocytes contained two GRCs, either a micro-GRC and a macro-GRC or else two micro-GRCs (Sotelo-Muñoz et al. 2022). This suggests that significant variation in the GRC size may exist not only between species but also within species and even within a single individual. Sotelo-Muñoz et al. (2022) suggested several mechanisms which could explain the origin of within-species polymorphism in GRC size. For example, fragmentation of the GRC during its elimination from the spermatocytes followed by paternal inheritance of the GRC fragment can lead to the origin of a smaller GRC in a population. A shorter GRC might also be the result of GRC fragmentation and loss of its parts during germline mitotic divisions. This sort of mutation would normally not be tolerated by the cell in standard A chromosomes, but given the enormous variability in GRC size even among closely related species, it is possible that large parts of this chromosome are in fact non-essential and thus their loss might not have large effects on their carrier’s fitness. At the same time, additions of new sequences to the GRC from standard chromosomes may be well tolerated as the presence of the GRC only in the germline reduces the pleiotropic effects of such mutations. It is plausible that once polymorphism in the GRC size exists in the population, occasional inheritance of two GRCs of different sizes by a single zygote and unstable mitotic inheritance of these GRCs may result in the observed GRC size mosaicism.
Currently, the frequency of polymorphism and mosaicism for GRC copy number and size in songbirds is difficult to estimate. Most species have only had a few individuals analyzed, and most of these individuals were males, making estimates for females especially uncertain. However, the data obtained to date indicates that polymorphism and mosaicism in GRC copy number could be relatively frequent across songbird species (Table 1).
Female meiotic drive and maternal inheritance of two GRC copies
An alternative explanation for where the two GRCs in females come from, as well as why polymorphism and mosaicism for GRC number occur, was proposed by Malinovskaya et al. (2020a). They suggested that zygotes of both males and females can already contain two GRC copies. Both copies would be inherited from the mother due to nondisjunction of GRC homologs in the first meiotic division (MI) and their preferential segregation into the egg (i.e., meiotic drive). During germline development, germ cells can actively eject or passively lose one of the GRCs. Since male germ cells undergo a much higher number of mitotic divisions before entering meiosis, they would be more likely to lose one of the GRCs and contain a single copy in the pachytene cells. This scenario does not exclude the possibility that zygotes with one GRC copy occasionally arise via normal segregation of two GRCs in female meiosis I. A single GRC could also be inherited from mothers carrying a single GRC in their pachytene cells, which would explain why some females have only one GRC.
This scenario could potentially explain sex differences, polymorphism, and mosaicism in GRC copy number. However, it depends on the validity of its key assumption: meiotic drive via nondisjunction and preferential segregation of both GRC homologs into the secondary oocyte (and then to the egg cell after normal segregation in meiosis II). In birds, the only known asymmetric divisions, which could provide a high efficiency of GRC accumulation, occur during female meiosis. Both polar bodies are formed at the periphery of the oocyte; therefore, if GRC homologs do not separate, they have a high chance to remain in the egg. Indeed, meiotic drive of B chromosomes during asymmetrical MI has been documented in females of many non-avian species (Hewitt 1976; Nur 1977; Nur and Brett 1985; Cano and Santos 1989; Santos et al. 1993). However, these studies mostly described the drive of a single B chromosome, which formed a univalent. Meiotic drive of GRC in MI requires nondisjunction of properly synapsed GRC bivalents. Malinovskaya et al. (2020a) suggested that nondisjunction can be facilitated by the extreme polarization of chiasmata positions in GRC bivalents. In females of all three species carrying macro-GRCs studied to date (zebra finch, sand martin, and great tit), recombination occurs in one or both ends of GRC bivalents (Fig. 1b; Pigozzi and Solari 2005; Malinovskaya et al. 2020a; Torgasheva et al. 2021). Such a polarized distribution of chiasmata is associated with an increased frequency of chromosome nondisjunction at the first meiotic division in other organisms (Sears et al. 1995; Koehler et al. 1996; Hassold and Hunt 2001).
However, recent observations reporting the lack of heterozygosity in zebra finch male siblings, which share the same GRC haplotype as their uncle from their maternal side (Pei et al. 2022), contradict the assumption that two homologous GRCs are transmitted to the progeny. They must have accumulated noticeable differences if passed through many generations and recombined in limited regions. Nevertheless, female meiotic drive at MI and inheritance of two GRCs from females could possibly occur at least in some species or individuals.
Zygotes with two GRCs can also occur via non-disjunction of GRC sister chromatids and their preferential segregation to the egg cell in the second meiotic division (MII; Fig. 4). Meiotic drive in MII, although less intuitive than in MI, can also occur due to the asymmetrical geometry of this division (reviewed in Clark and Akera 2021). One may speculate that meiosis-specific cohesins or other meiotic players controlling the correct separation of sister chromatids or centromeres might be involved in the GRC nondisjunction at MII in a similar way as has been described for B chromosome drive during the first pollen mitotic division (Ruban et al. 2020). Occasional normal disjunction in MII and rare nondisjunction during premeiotic mitoses can explain how polymorphism and mosaicism for GRC copy number arise in females. In addition, under this scenario even females with a single GRC would produce gametes with two GRC copies, thereby maintaining the polymorphism in GRC copy number in the population.
GRC elimination from somatic cells and male germ cells
Another question is how the GRC is eliminated from somatic cells during early embryogenesis and from male germ cells in spermatogenesis. Currently, the mechanisms of GRC elimination from somatic cells remain entirely unknown. Pigozzi and Solari (2005) hypothesized that GRC elimination from somatic cells occurs via different mechanisms in males and females. In females, nondisjunction of the GRC chromatids and their segregation to a germline progenitor cell would leave a somatic progenitor cell without a GRC. In males, lagging of one of the GRC chromatids during mitotic anaphase would lead to a somatic progenitor cell without a GRC and a germline progenitor cell with a single GRC (Fig. S1). Such sex-specific differences in GRC behavior would, however, require sex differences in gene expression already at early stages of embryo development when the germline is determined. Alternatively, the GRC might be epigenetically modified in both sexes and marked for elimination from somatic cells in a similar way as it has been observed in spermatogenesis (del Priore and Pigozzi 2014; Malinovskaya et al. 2020a). Cytological observations of the earliest stages of songbird embryonic development are needed to shed light on the details of the GRC elimination from somatic cells.
A few pilot studies on mechanisms of GRC elimination from male germ cells during spermatogenesis have already been published (del Priore and Pigozzi 2014; Goday and Pigozzi 2010; Malinovskaya et al. 2020a; Schoenmakers et al. 2010). They showed that from the very beginning of meiotic prophase, the single GRC is heterochromatic in primary spermatocytes, marked with specific histone modifications during prophase (e.g., H3K9me3, H3K9me2, and MacroH2A), and shifted to the nuclear periphery. The GRC is then observed in the cytoplasm of secondary spermatocytes, suggesting that its elimination from the nucleus occurs during the first meiotic division. Later, the GRC is seen as a micronucleus in the cytoplasm of secondary spermatocytes and young spermatids, and finally expelled from the cells. A similar mechanism might be involved in GRC elimination from male germ cells during their pre-meiotic mitotic division if a zygote receives two GRC copies (see above). This is supported by cytological observations of pale martin spermatogonia containing two GRC copies, one of which is located within the nucleus and the other one is moved to the cell periphery and almost expelled (Fig. 5c in Malinovskaya et al. 2020a).