CRISPR Fish Reel in Novel Roles for Estrogen Receptors in Reproduction

The physiologic effects of estrogenic steroids are controlled primarily via transcription factors in the nuclear receptor superfamily, namely the estrogen receptors. Whereas two estrogen receptor (ER) genes are encoded in mammalian species, ESR1 (ERα) and ESR2 (ERβ), vertebrate species from the teleost lineage (bony fish) possess a single ERα gene (esr1) and two ERβ genes, ERβ2 (esr2a) and ERβ1 (esr2b). The two ERβ genes are thought to be products of whole genome duplication events during ancient teleost evolution, and retention of duplicated genes over evolutionary time implicates distinct functions of these paralogs in physiologic processes. However, the actions of ERs have not been closely examined past the embryonic or larval stages of the zebrafish; this contrasts with the extensive data from genetic mouse knockouts of ERα and/or ERβ (ERKO) (1, 2). This has left a critical gap in understanding of the postnatal functions of zebrafish ERs. In this issue, Lu et al. (3) report on genetic knockout of esr1, esr2a, and esr2b in zebrafish using CRISPR/Cas9 genome editing technology and explore for the first time the functions of the three ER subtypes in adult zebrafish.

Initial cloning of the esr1, esr2a, and esr2b genes by Menuet et al. (4) indicate differential E2-dependent regulation and transcriptional activity among the three zebrafish ERs, as well as distinct expression patterns. Prior to the current study, morpholino-mediated knockdown of zebrafish ERs and ER subtype-selective antagonists were used to demonstrate subtype-specific roles of ERs during early embryonic development. In particular, Hu et al. (5) showed that selective antagonism of ERα and ERβ with small molecule inhibitors, as well as knockdown of esr2a and esr2b, caused diverging phenotypes in primordial germ cells, which precede gonadal cells. These observations in primordial germ cell phenotypes bring into question the distinct functions of the ERα and ERβ1/β2 subtypes in zebrafish gonad development and eventual fertility, which have evaded direct study because of the transient nature of the tools available previously. With the stable knockout of ERs via CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 technology, Lu et al. (3) are now able to shed light on ER’s reproductive roles in zebrafish.

Sex determination and subsequent fertility in fish make up a highly flexible process that can be controlled by multiple genes as well as exogenous factors; sex steroid signaling may provide a powerful stimulus to control gonadal differentiation. Strikingly, Lu et al. (3) show that genetic knockout of a single ER subtype alone did not interfere with male or female gonadal development and fertility. Rather, the group observed striking intersexual gonad phenotypes develop at 60 days post fertilization (dpf) in female esr2a^−/−;esr2b^−/− and esr1^−/−;esr2a^−/−;esr2b^−/− knockout fish. Remarkably, as early as 45 dpf, these mutant females began to exhibit follicular degeneration and by 60 dpf exhibit ovo-testes gonads or complete sex reversal. These phenotypes suggest a cooperative role of all ER subtypes in zebrafish female sex maintenance in addition to gonadal differentiation. Although ERs were critical in female gonadal maintenance in this study, the zebrafish single and compound ER mutant males and females bred normally and produced healthy progeny, even after sex reversal. These observations are in stark contrast to ERKO mice, which exhibit defects in fertility. Mature αERKO mice of both sexes are infertile, but among βERKO mice, only females are subfertile. Double knockout (αβERKO) are viable but infertile in both sexes. Although the requirements for ER in fertility differ between mice and zebrafish, mouse αβERKO females exhibit intersexual gonads, which echo intersexuality phenotypes observed in double-knockout esr2a^−/−;esr2b^−/− and triple knockout esr1^−/−;esr2a^−/−; esr2b^−/− zebrafish mutants. Taken together, the observations made in these zebrafish knockouts may enable a previously unappreciated opportunity to study novel functions of ERs, which may have otherwise been masked by fertility phenotypes in mouse models.

In addition to the current study of ER knockout zebrafish lines, Lau et al. (6) recently reported an analogous study of the effects of aromatase (cyp19a1a) knockout on gonadal development. Because aromatase knockout will eliminate production of estrogenic steroids, it may be expected to phenocopy combined ER knockouts. However, the phenotypes were in fact distinct, with aromatase knockout halting ovary development at an earlier stage (differentiation of the juvenile ovary) relative to ER knockouts (after ovarian differentiation, during folliculogenesis). Although the authors ascribe this difference to a role for extranuclear ERs, another potential factor in modulating ovarian development may be androgen receptor (AR) signaling. Androgenic steroids increase in patients with breast cancer receiving aromatase inhibitors, and aromatase-deficient humans and mice also have elevated androgens (7). This suggests that aromatase-knockout zebrafish may indeed have elevated androgens and AR activity—without concurrent ER activity—throughout development. Although hyperandrogenic conditions are detrimental to ovarian function and fertility in adult mice, AR is also required for normal follicle development (8). This complicated interplay between AR and the broader ER family may indeed now be best explored in zebrafish using the powerful CRISPR/Cas9 systems described by Lu et al. (3).

Altogether, the development of a bona fide genetic ER knockout zebrafish may provide an opportunity to map critical functions of ERs, both within the scope of zebrafish reproductive development and in relation to murine ERKO models. For the former, expression analysis of each ER in specific cell types (as discussed in the current study) provides the opportunity to identify potential cell-specific ER subtype functions or to explain observed redundancy in ER functions. In particular, the relative requirement for esr2a suggests that esr2a may have either unique, critical transcriptional targets or unique factors regulating its expression/function that should be identified. For the latter, the dispensable state of esr1 in zebrafish fertility is in stark contrast to the severe effects of Esr1 knockout in mice. Understanding the differences in regulation, transcriptional targets, genomic binding, and other functions of ERs using the respective knockout systems may provide unique insight on critical functions of ERs in human physiology and pathology.