Introduction
Naturally occurring sexual shifts during hermaphrodite life cycles offer unique opportunities to study sexual plasticity and sexual differentiation. The latter processes are currently attracting worldwide interest due to accumulating evidence that environmental factors and endocrine-disrupting chemicals affect hormonal regulation, sexual development and fertility in animals (Olmstead and LeBlanc, 2000; Rodriguez et al., 2000, 2007; Ford et al., 2003, 2004; Hayes et al., 2010; Ford, 2012).
Important examples of sexual plasticity may be found among the crustaceans (Pandian, 2016), an ancient highly diverse group of animals in which a variety of reproductive strategies are represented, including gonochorism (separate sexes) (Juchault, 1999), intersexuality (combination of male and female features within a gonochoristic species) (Goldschmidt, 1938; Reinboth, 1975; Sagi et al., 1996; Abdel-Moneim et al., 2015; Levy et al., 2020c), asexual reproduction (e.g., parthenogenesis) (Scholtz et al., 2003; Martin et al., 2007), and different types of hermaphroditism (Levy et al., 2018; Benvenuto and Weeks, 2020). The last of these strategies constitutes a means of reproduction in which a particular individual bears both ovarian and testicular tissues [ovotestis (Hoffman, 1972; Stentiford, 2012)] and produces gametes of both sexes (Benvenuto and Weeks, 2020). Hermaphroditism may be either simultaneous, in which the individual functions both as a male and a female (Bauer and Holt, 1998; Baeza, 2007), or sequential, in which the individual first matures as one sex and then transforms irreversibly into the other (Hoffman, 1968; Subramoniam, 1981; de Almeida and Buckup, 2000). Of relevance to this study, hermaphroditism may be exploited as a valuable model system for studying sexual plasticity, since the sex-differentiation process does not occur as a single event limited to the early developmental stages of the organism.
A fairly well researched case study of hermaphroditism in crustaceans is that of the shrimp Hippolyte inermis, formerly known as H. viridis (Reverberi, 1950), a caridean protandric species inhabiting seagrass (Posidonia oceanica) meadows in shallow waters of the Mediterranean Sea and the Atlantic coasts of Spain (Zupo and Messina, 2007). While protandric species are commonly born as males, followed by a transitional stage before transforming to females (Yaldwyn, 1966), H. inermis diverts from this pattern in that it lacks the ovotestis-containing transitional stage (Cobos et al., 2005; Zupo and Messina, 2007; Mutalipassi et al., 2018) in both of its reproductive cycles, one in the spring and the other in the fall (Figure 1). In the spring, these shrimp exhibit a first reproductive burst, and approximately three months after hatching, both immature males and females are present in the population. In contrast, at the end of the second reproductive burst in the fall, only males are present; these animals transform into females in the following spring (Zupo, 1994). It has been suggested that the spring sexual shift in H. inermis is promoted by a diet based on diatoms (Zupo, 2000), namely, the species of Cocconeis, that are abundant in their seagrass habitat, resulting in an early transformation of males at a younger age and, hence, in the presence of small females in the population by the end of the spring (Zupo, 1994, 2000). However, the lower abundance of these diatoms in the fall leads to a normal process of protandric development, with a predominance of young males in the population following this second reproductive burst. Physiologically, it has been suggested that compounds present in the Cocconeis spp. diatoms cause apoptosis of the crustacean androgenic gland (AG) and, consequently, control the sexual shift from maleness to femaleness (Zupo et al., 2007). This suggestion is in keeping with the universal mechanism of control of crustacean sexual differentiation by the insulin-like androgenic hormone (IAG) – secreted by the AG – in a process involving many upstream and downstream genes, termed the “IAG-switch” (Levy and Sagi, 2020). When the switch is “turned on” (i.e., in normal males or through activation of the AG and expression or induction of IAG), masculinization occurs, whereas when it is “turned off” (i.e., in normal females or through AG ablation or silencing or inactivation of the IAG), the result is feminization (Charniaux-Cotton, 1958, 1962; Nagamine et al., 1980a,b; Ventura et al., 2012; Levy et al., 2016).
Another case study of crustacean hermaphroditism – one that we also exploit in this study – is that of the protandric Northern spot shrimp, Pandalus platyceros, which is widely distributed in the North Pacific Ocean (Butler, 1964). Unlike H. inermis, in which the transformation from maleness to femaleness takes place up to few months following hatching (Zupo, 1994), the transformation in P. platyceros is slower, occurring at the age of 3 to 5 years (Butler, 1965; King and Moffitt, 1984; Iversen et al., 1993; Kimker et al., 1996). Therefore, unlike gonochoristic species in which the IAG-switch-based sexual differentiation process is limited to early developmental stages, protandric shrimps, such as H. inermis and P. platyceros, may serve as models in the study of sex-controlling toolkits, because in such species the process is not confined to early development but rather occurs later in the life cycle, when adult males transform into females. Protandry may thus be regarded as offering an opportunity to obtain new insights into sexual differentiation that cannot be obtained from studies of gonochoristic species.
In the present study, we collected samples of H. inermis at different stages of the species’ protandric life cycle and constructed RNA libraries that yielded both known and novel, yet unannotated, sex-specific genes that are assumed to be associated with sexual differentiation and reproduction. To obtain insight into the function of these novel genes, we also performed a comparative in silico analysis with P. platyceros, taking advantage of RNA libraries of specific tissues previously obtained over the life cycle of P. platyceros (Levy et al., 2020a). P. platyceros is particularly suitable for such an analysis for two reasons: (i) taxonomically, the two species, H. inermis and P. platyceros, belong to families – Pandalidae and Hippolytidae, respectively (Martin et al., 2009) – that are closely related within the infraorder Caridea (Christoffersen, 1990; Wolfe et al., 2019; Levy et al., 2020c), and (ii) technically, the RNA for the H. inermis libraries was extracted from the entire body of the animal (due to its small size that prevented dissection of specific tissues), while the RNA for constructing P. platyceros libraries was obtained by dissecting out specific tissues (Levy et al., 2020a). Therefore, our working hypothesis for the in silico analysis part of this study was that a search for homologs to the newly discovered H. inermis genes in transcriptomic libraries of P. platyceros (and other decapod crustaceans) might reveal the function of those genes in a more general context that includes a wider range of crustaceans.