Inhibition of ALDH1A (RALDH) leading to impaired fertility via disrupted meiotic initiation of fetal oogonia of the ovary

• Terje Svingen
• Cataia Ives

• You Song

Revision dates for related pages

• Sex: This AOP applies to females. atRA is also involved in meiosis of testicular gonocytes, but this occurs postnatally. In the female ovaries, atRA induces meiosis of oocytes during gestation, thus the spatiotemporal expression of atRA in the ovaries are tightly controlled. Finally, as this AOP is concerned with establishing the ovarian reserve/follicle pool through mechanisms that are unique to ovaries, restricting the AOP to female only is appropriate.
• Life stages: This AOP spans the period from mid- to late-gestation in mammals, all the way to adulthood where fertility is manifested. The upstream event pertains to fetal/neonatal life stages, whereas the downstream events pertain to adult reproductive life stages. 
• Taxonomy: Strongest evidence for the role of atRA in regulating oocyte entry into meiosis stems from moue studies, so the taxonomic applicability is strongest for this animal model. Evidence also exists for the same modes of action being relevant across mammalian species, including human (Kalampokas et al, 2014), albeit the evidence for taxonomic applicability is still weaker.

The critical first stage of this AOP is the lack of atRA in ovaries at the stage where oocytes need to enter meiosis during gestation. Failure to enter meiosis at the correct time during development is detrimental to oocyte development and ultimately this will compromise the follicle pool in adulthood; a non-renewable source for producing viable eggs for fertilization. However, reduced atRA is not defined as the MIE in this AOP, as atRA synthesis by the action of ALDH1A enzymes is required, and this synthesis step is a potential vulnerable point for chemical perturbation. 

• MIE (Inhibition, ALDH1A): atRA is synthesized from dietary vitamin A in a two-step enzymatic process, where ALDH1A is responsible for the second, irreversible oxidation of retinal to atRA (Chatzi et al, 2013; Shannon et al, 2017).
• KE1 (Decreased atRA): Although ALDH1A enzymes can make different retinoid metabolites (9-cis, 13-cis, all-trans), atRA is considered the only active metabolite in mammals (Krężel et al, 2019). atRA acts locally in tissues as a paracrine, short-range, signaling molecule (Teletin et al, 2017). In the mouse, atRA is synthesized in the adjacent mesonephros and diffuses into the gonad proper to establish local concentrations required for organogenesis (Chassot et al, 2020; Kumar et al, 2011), including prompting oocytes to enter meiosis (Bowles et al, 2006; Koubova et al, 2006). In humans, atRA appears to be synthesized locally in the ovary. This KE is essential for this AOP as represents the essential link between ALDH1A inhibition by chemicals with reduced fertility in adult females.
• KE2 (disrupted meiosis): The downstream KE of this AOP, reduced follicle pool, can arise from multiple events, of which disrupted meiosis is only one. Nevertheless, disrupted meiosis is essential for this AOP in that it is the clear rational link between inhibited retinoid signaling and reduced fertility via diminished ovarian reserve. For meiosis to initiate in the mouse, oocytes need to express Stra8 (Baltus et al, 2006), a factor that is regulated by atRA via retinoid receptors RAR/RXR (Bowles et al, 2016; Bowles et al, 2006; Feng et al, 2021; Koubova et al, 2006). Although the role for atRA for initiating meiosis in humans is still under some debate, the role for Stra8 appears essential also for human oocytes to initiate meiosis (Childs et al, 2011).
• KE3 (reduced ovarian reserve): In mammals, it is broadly accepted that females are born with a set number of follicles, termed the ovarian reserve, which is dependent on proper development, including meiotic initiation, during fetal life (Grive & Freiman, 2015). A large number of oocytes are lost during ovarian/oocyte development to ensure a quality ovarian reserve. However, a minimum amount of oocytes are required to establish and maintain adult reproductive function. Therefore this KE represents an essential step in the AOP, since falling below a critical number of follicles will lead to disrupted ovary function and irregular cyclicity (as normally occurs during menopause).
• KE4 (irregular ovarian cycle): Female fertility depends on the ovarian cycle to produce competent follicles for ovulation and fertilization. Thus, this KE is an essential step in determining fertility status.  
• AO impaired fertility, female: Fertility represents the capability to reproduce and as such is the essential AO of this AOP. It is measurable both at the individual and population level. Although the AO of this AOP (ID 406) describe impaired fertility independent of sex, the AOP is specific to females as it involves oocyte/ovary development and function.  

Biological Plausibility, coherence, and consistency of the experimental evidence

The role for ALDH1A2 in the synthesis of atRA is well established as an essential component of regulating regional expression of retinoid species during development. It is also well established that atRA is an inducer of meiosis in germ cells in mice; however, there is some debate about the essentiality of atRA in this process in human fetal ovaries. The requirement for oocytes to enter the first phase of meiosis during fetal development is also well established, hence the biological plausibility linking meiotic failure with loss of oocytes at later developmental stages is strong.

Although non-meiotic oocytes can survive in germ cell nests and during nest breakdown, they will ultimately be eliminated from the oocyte pool of competent follicles. There is therefore a direct link between meiotic entry and fertility during adulthood. Thus, this AOP provides a plausible chain of events linking reduced atRA during fetal life with reduced ovarian reserve and fertility during reproductive age. The strength of the downstream KEs and KER – reduced ovarian reserve and reduced fertility – is very well documented and thus the biological plausibility is very strong. Evidence for a direct link between the AO and perturbed atRA synthesis, or reduced atRA levels, during early development comes mainly from mouse studies; yet the relationship is regarded biologically plausible also in humans, but with weight of evidence not being as strong. 

Concordance of dose-response relationships

The quantitative understanding of dose-response relationships in this AOP is limited. Whilst the relative levels of endogenous atRA produced by the ovary (for any species) remains unknown, similarly, the quantitative relationship between atRA levels and induction of meiosis also remains unclear. Nevertheless, it is has been conclusively shown that low levels of exogenous atRA can induce mouse and rat germ cells to enter meiosis both in vitro and ex vivo (Bowles et al, 2006; Livera et al, 2000). Likewise, atRA is necessary to achieve meiosis in in vitro-derived oocytes via PGCLCs (Miyauchi et al, 2017).

Temporal concordance among the key events and the adverse outcome

This AOP bridges two different life stages: fetal/perinatal and adult/reproductive age. The adverse outcome is the result of perturbation taking place during early stages of ovary development. In mice, rats and humans, the oocytes must enter meiosis prophase in order to establish the follicle pool/ovarian reserve postnatally. Thus, the AOP focusses on chemical perturbations during fetal life, which occurs around E13-E16 in mice and E15-E18 is rats, or first trimester in humans (Peters, 1970), but the adverse outcome does not manifest until adulthood. 

There is strong temporal concordance between the various key events, from inhibition of ALDH1A2 (RALDH2) that leads to reduced atRA synthesis. In turn, atRA must be present in the fetal ovaries at the time when oocytes are supposed to enter meiosis mid-gestation in mice (or first trimester in human). With a significant reduction in available atRA the oocytes will not enter meiosis, ultimately leading to the downstream key event of loss of oocytes beyond what is normal. The number of oocytes, or the oocyte pool/ovarian reserve, in turn will affect ovary function and fertility at reproductive stages, hence the temporal sequence of events is rational based on the biological process.

Strength, consistency, and specificity of association of adverse effect and initiating event

In mice, there is strong evidence to support the view that atRA is an inducer of meiosis in germ cells, with consistent results from in vitro (PGCLCs), ex vivo (ovary cultures) and in vivo studies as listed under KE 2477. There is strong evidence showing the importance of RA for female fertility, but this relates to many aspects of reproductive development and function from fetal life to adulthood, including maintaining pregnancy (Clagett-Dame & Knutson, 2011). Thus, it can be difficult to distill exactly how atRA-controlled meiotic entry of oocytes directly link to reduced fertility. Nevertheless, a direct relationship is strongly supported by the fact that Stra8-depleted mice are infertile with small ovaries lacking oocytes (Baltus et al, 2006) and that Stra8 induction in germ cells is controlled by atRA in mice, rats and humans (Bowles et al, 2006; Childs et al, 2011; Koubova et al, 2006; Livera et al, 2000). Furthermore, vitamin A-deficient (VAD) mice display delayed or failed meiotic entry of fetal oocytes depending on level of Vitamin A deficiency (Li & Clagett-Dame, 2009).

Uncertainties, inconsistencies and data gaps

In mice, there is strong evidence to support the view that atRA is important for initiating meiosis in germ cells (Bowles et al, 2016; Spiller et al, 2017; Teletin et al, 2017). Some studies suggest that atAR is not critical but important for meiotic entry under normal physiological conditions by evidencing meiosis in Aldh1a1, Aldh1a2 and Aldh1a3 ablated mice, individually and in tandem (Bellutti et al, 2019; Chassot et al, 2020; Kumar et al, 2011); however, additional studies have shown redundant roles between all three Aldha isoforms which can compensate for deletion of one or two (Bowles et al, 2016). More specifically, both double (Aldh1a2/3) and triple (Aldh1a1/2/3) knockout mouse models display reduced Stra8 expression in oocytes, yet oocytes eventually go through meiosis, which could suggest a redundant role for atRA for meiosis in the ovaries (Chassot et al, 2020; Kumar et al, 2011). A similar phenotype with reduced Stra8 expression but eventual meiotic initiation is seen for deletion of atRA receptors RAR-α, -β, -γ) in mice (Vernet et al, 2020). But, although RAR knockouts were also capable of producing offspring, it remains unclear if any of the above-mentioned mouse models display impaired fertility or whether the size of their oocyte pools are affected.

This AOP is still largely qualitative, as the quantitative understanding between chemical potency and perturbation of KEs are insufficient. This relates to the dose-response relationship between concentrations of atRA in the ovary relative to meiotic initiation of oocytes. It also relates to the relationship between number of lost oocytes during development relative to the oocyte pool/ovarian reserve, as there naturally is a large loss of oocytes during development.