LAUSR

It has long been known that female and male germ cells can be isolated from fetal gonads and coaxed, in vitro, to enter meiosis—the hallmark of gametogenesis (Handel et al., 2014). However, progression through meiosis could not be faithfully replicated outside the gonad. It seemed almost inconceivable that germ cells could acquire the morphological and molecular features of mature germ cells without also having completed meiosis. Thus, the idea took hold that gametogenesis in vivo is such a complex process that it could not possibly be completed correctly in vitro—a seemingly insurmountable obstacle. Had this idea held sway over experimental science, we most probably would not have written this editorial to promote this call for manuscripts. This dogma began to be challenged in 2003, with studies that showed morphogenesis of mouse embryonic stem cells into oocytes in vitro (Hubner et al., 2003), into spermatozoa in vivo (Toyooka et al., 2003) and haploidization in vitro (Geijsen et al., 2004; Nayernia et al., 2006). Clearly, these haploid cells had progressed through meiosis, as determined by flow-cytometric analysis of DNA content without karyotypization. However, while some stages of gametogenesis that occur in the fetus could be reproduced in vitro, the full course of gametogenesis still required transplantation of germ cell precursors in the natural environment of the gonads. While this is generally unproblematic in animal models, it would pose a limitation in humans. In the years following 2003, increasing numbers of investigators succeeded in reducing the necessity for in vivo transplantation, by refining in vitro systems that support gametogenesis, and the most recent studies show that meiosis in spermatogenesis (Zhou et al., 2016) and oogenesis (Hikabe et al., 2016) can be achieved entirely in vitro from pluripotent cell lines. The in vitro-generated mouse germ cells were functional; they differentiated into spermatozoa or oocytes that gave rise to healthy offspring. Of course this feat transcends the generation of mice: it establishes that the Weissman barrier can be broken, not only by returning a somatic nucleus into an oocyte (cloning by somatic cell nuclear transfer (SCNT)) but also by producing the oocyte itself (Sabour and Scholer, 2012). Why is this so noteworthy? It means that, little by little, we are taking full control of the mammalian germline, starting in the mouse, which is often the first step toward achieving the same result in other species including humans. What changes, then, with the possibility of generating gametes in vitro from natural precursors and from reprogrammed cell lines? Future developments will range from scientific to clinical. The following outline of a small selection of a large body of recent publications aims to demonstrate what impact in vitro gametogenesis can offer. In basic science, the creation of in vitro differentiation systems has often heralded the characterization of the signaling pathways involved in the natural counterpart. What are the signals that underlie germ cell proliferation, growth and differentiation? The culture conditions that support in vitro gametogenesis contain growth factors and hormones. In vivo, the breakdown of mouse germ cell cysts in the genital ridge provides a quality check and that the prevalence of apoptosis is high (Pepling and Spradling, 1998). When mouse primordial germ cells are pushed through oogenesis in vitro, cysts break down prematurely and this comes with a massive loss of oocytes, unless an estrogen receptor antagonist is added to the in vitro culture to put a brake on the cyst breakdown (Morohaku et al., 2016). These cysts are considered to be a hallmark of mammalian germ cell differentiation. Recently, these results have been reproduced with reprogrammed cell lines (iPS cells) (Hikabe et al., 2016). This may suggest that we are compensating for some later restriction by getting more germ cells through the cyst stage, only to let them fail later, a hypothesis supported by the observation that the majority of in vitro-produced oocytes fail at the developmental test of totipotency (Hikabe et al., 2016). There are likely intrinsic differences between gametes produced in vitro and natural gametes, because the mechanisms—while similar—are unlikely to be exactly the same. Indeed, the behavior of the in vitro-produced mouse gametes presents some notable differences from their in vivo counterparts. Thus far, the in vitro-produced mouse gametes appear to be functionally inferior to natural gametes, reminiscent of the comparison between embryonic stem cells derived from fertilized or cloned embryos and those derived without oocytes via iPS cell technology (Colman and Burley, 2014). Thus, although functional gametes can be produced in vitro, in vivo gametogenesis remains superior, for as-yet undefined reasons. One possible candidate may be genetic imprinting. Despite major concerns that any culture system is suboptimal when it comes to preserving the epigenetic marks, Hikabe et al. (2016) showed us