RXR Expression in Marine Gastropods with Different Sensitivity to Imposex Development

The superposition of male sexual characteristics in female marine gastropods (imposex) represents one of the clearest ecological examples of organotin-mediated endocrine disruption. Recent evidences suggest that signaling pathways mediated by members of the nuclear receptor superfamily, RXR and PPARγ, are involved in the development of this pseudohermaphroditic condition. Here, we identified significant differences in RXR expression in two caenogastropod species from Nuevo Gulf, Argentina, Buccinanops globulosus and Trophon geversianus, which present clear contrast in imposex incidence. In addition, B. globulosus males from a polluted and an unpolluted area showed differences in RXR expression. Conversely, PPARγ levels were similar between both analyzed species. These findings indicate specie-specific RXR and PPARγ expression, suggesting a major role of RXR in the induction of imposex.

It has been shown that TBT binds RXR and activates the RXR-peroxisome proliferator-activated receptor gamma (PPARγ) heterodimer 33 . Direct administration of 9-cis retinoic acid (9cRA) induced imposex in T. clavigera 18,34 and N. lapillus 14 . However, the application of rosiglitazone (Rosi), a PPARγ ligand, to N. lapillus, induced imposex in the absence of TBT 35 . All these data suggests the direct association between TBT-RXR-PPARγ with imposex development.
The nassarid gastropod Buccinanops globulosus (Kiener, 1834) inhabits Nuevo Gulf and is distributed along the Argentinean coasts in sandy bottoms of shallow waters 36 . The genus Buccinanops has been affected by imposex in Argentina 37,38 . Trophon geversianus (Pallas, 1774) is a muricid gastropod widely distributed along the Southern Atlantic and Pacific coasts [39][40][41] . The specie is present in subtidal and intertidal habitats from Nuevo Gulf. Imposex incidence is very low in T. geversianus, even in harbor areas with TBT contamination. In contrast, B. globulosus presents high sensitivity to imposex incidence even at low environmental TBT concentrations 41 .
Working with species that inhabit the same area being exposed to the same environmental conditions but show physiological differences in the development of imposex, will allow us to expand our knowledge on the metabolic pathways altered by organotin compounds. Thus, the specific aim of this study was to evaluate the expression of RXR and PPARγ in two caenogastropods from Nuevo Gulf, Argentina, B. globulosus and T. geversianus, to explain some aspects of its differential sensibility to imposex development.

Materials and Methods
Collection of animals. Sexually mature males and females of B. globulosus and T. geversianus were collected during July-August 2016 from two places in Nuevo Gulf, with the same physical conditions and separately by 20 km, Cerro Avanzado beach (CA), a zone with scarce maritime traffic and very low imposex incidence, and from Luis Piedra Buena Harbor (LPBH), located in the city of Puerto Madryn 42 ( Supplementary Fig. S1). LPBH is an area with high maritime traffic characterized by the presence of commercial, fishing and recreational vessels. A total of 895 boats with up to 294 m length, arrived in the city in 2016/2017 season 43 . Aluminum derivates, porfids and others materials are transported during the year. High levels of trace metals, polycyclic aromatic hydrocarbons (PAHs) and TBT were previously recorded in sediments and organisms 41,44-48 . Tissue preparation. After removal of the shell, the percentage of imposex-affected females was calculated in terms of penis and vas deferent presence, and the relative penis length index (RPLI) was calculated 49 . Male penis and penis/penis-forming area from imposex-affected female were excised out and fixed in formalin solution (4% in PBS) or immediately frozen at −80 °C. Immunohistochemistry (IHC). Sample sections of formalin-fixed, paraffin-embedded tissues were stained with hematoxylin-eosin (H&E) or reacted with RXR (sc-774) and PPARγ (sc-7196) antibodies (Santa Cruz Biotechnology) using the avidin-biotin peroxidase complex technique (Vectastain Elite ABC kit; Vector Laboratories). Briefly, 3 penis sections from five independent individuals of both species were rehydrated from xylene to 70% ethanol passing through decreased graded ethanol, before endogenous peroxidase activity inhibition (10% H 2 O 2 in 70% ethanol). Antigen retrieval with HCl 2 N was performed before immunostaining. After PBS washes, sections were blocked (2.5% BSA in PBS) and incubated overnight at 4 °C with the primary antibody. After biotin-conjugated secondary antibody (incubation for 1 h at room temperature, the reaction was developed using the DAKO Liquid DAB + Substrate Chromogen System (K3468, DAKO) according to the manufacturer's protocol under microscopic control. Specimens were counterstained with hematoxylin, dehydrated and mounted. Positive cells were counted in 10 high-power fields (HPFs) of each section, using 1000× magnification, and expressed as the mean ± SEM of the percentage of the ratios between the number of events and the cell number/ HPF. Tissue extracts and western blot. Frozen tissues from five independent individuals of both species were homogenized in ice-cold TEDGS 10% buffer (50 mM Tris pH = 7.4, 7.5 mM EDTA, 0.5 mM dithiothreitol, 10% glycerol, 0.25 M sucrose), including protease inhibitors. The homogenate was centrifuged (20 min, 12000 rpm at 4 °C) and total protein concentration in each supernatant was determined by Lowry method 50 . Equivalent amounts of protein (100 μg) from tissue lysates were separated on discontinuous polyacrylamide gels and detected by western blot. Membranes were probed with RXR (sc-774), PPARγ (sc-7273) or β-Actin (sc-47778) antibodies (Santa Cruz Biotechnology) overnight at 4 °C, and then incubated with horse anti-mouse or goat anti-rabbit peroxidase-conjugated secondary antibody (Vector Laboratories). The luminescent signal was generated by enhanced chemiluminescence (ECL) method and the blots were exposed to a medical X-ray film (ortho CP-GU, AGFA). Band intensity was determined by densitometry using ImageJ 1.47 v software (https://imagej.nih. gov/ij/download.html) 51 .
Statistical analysis. ANOVA and the Tukey multiple post t test were used to analyze the differences of means of multiple samples; the Student's t test was used to compare the means of two different groups. In all graphs, the mean ± SEM is shown, and experiments were repeated at least three times. Significant differences between groups are indicated with asterisks (*p < 0.05; **p < 0.01; ***p < 0.001).

Imposex incidence in B. globulosus and T. geversianus. B. globulosus females collected from Luis
Piedra Buena Harbor (LPBH) presented 100% imposex incidence with different degrees of penis development (between 0.5 to 7.7 mm of length); meanwhile the opposite was observed for T. geversianus specimens (Table 1 and Fig. 1a). No imposex incidence was registered in female gastropods of both species collected in Cerro Avanzado beach (CA).

Evaluation of RXR and PPARγ expression. Immunohistochemical stains against RXR and PPARγ illus-
trate the localization of both proteins in the penis of B. globulosus and T. geversianus male gastropods collected from CA beach (Fig. 2a). RXR and PPARγ were detected in the nuclei of epithelial cells and smooth muscle cells surrounding the epithelium of vas deferens. No immunoreactivity was detected in the absence of primary antibodies (Fig. 2b), confirming binding specificity. www.nature.com/scientificreports www.nature.com/scientificreports/ We found a significant lower expression of RXR in penis of T. geversianus males in comparison with the same tissue of B. globulosus (a similar trend was found in epidermal cells of male penis, and in female gonads and digestive glands, Supplementary Fig. S2). In the case of PPARγ, no differences were observed between both analyzed species (Fig. 2c). These results were validated by western blot (Fig. 2d).
To further evaluate the expression of RXR and PPARγ in gastropods from a polluted (LPBH) or unpolluted (CA beach) areas, we used penis tissues from imposex-affected females and males of B. globulosus. We observed RXR and PPARγ expression in males from LPBH and CA beach and in imposex-affected B. globulosus females collected at LPBH (Fig. 3). Significant differences in RXR penis tissues expression were observed among males from CA and LPBH, and males from CA and imposex-affected females from LPBH, while no differences in PPARγ expression was detected (Fig. 3b,c).

Discussion
In this study, we compared for the first time the expression of RXR and PPARγ in two caenogastropod species from South Atlantic, which present clear contrasts in imposex incidence. Most of biochemical or molecular studies related to imposex have been performed under experimental laboratory conditions. Ours, is an entirely field work reflecting what occurs in natural conditions underscoring the relevance of the data. We found that penis of B. globulosus males express higher levels of RXR than the same tissue of T. geversianus males, collected in an unpolluted area (CA) of Nuevo Gulf. These results suggest specie-specific differences in RXR expression that could explain differences registered in penis length between both species (and probably in other neogastropods), and may also be related to their different biological response to environmental contaminants. As was previously mentioned, the hypothesis about the abnormal modulation of RXR as a mechanism by which organotins induce imposex in caenogastropods would turn out to be the most accepted one 20,52,53 . Therefore, RXR could be a central factor in the molecular mechanism involved in differentiation, proliferation, and morphogenesis of penis in male and imposex-affected female gastropods. Nevertheless, the low expression of RXR in penis from T. geversianus males, relative to that in B. globulosus males, is not an impediment to the normal development of secondary sexual characters as the penis. In addition to the local action that RXR would have in penis development, which would be under the control of the central nervous system (CNS) of gastropods, CNS was described as a tissue with high RXR expression able of being retinoid target 21 . It was proposed that a "penis morphogenetic factor (PMF)" Trophon geversianus (T.g.) males collected in CA. β-Actin was used as a loading control. The band intensity ratios of RXR and PPARγ expression relative to β-Actin were plotted (right, ***p < 0.001). Mouse lung and kidney tissues were used as positive controls. Full-length blots are presented in Supplementary Fig. S3. www.nature.com/scientificreports www.nature.com/scientificreports/ released by CNS under the control of RXR signaling is induced in males during penis differentiation 21 . That may explain the normal development of T. geversianus penis even having low levels of RXR expression, relative to B. globulosus males. However, this hypothesis and the identification of a PMF in T. geversianus and other gastropods need to be addressed. Future proteomic strategies, comparing both species under study, will shed light on further important factors or signaling pathways involved in differentiation, growth and penis formation in males.
Interestingly, no differences were observed in PPARγ expression between the species studied, highlighting the role of RXR in normal or abnormal physiological processes. Convergence of 9cRA and PPAR signaling pathways through PPAR/RXRα heterodimerization is well established in mammals 54 , a crosstalk that is recently starting to be explored in gastropod mollusks affected by imposex. Two PPAR homologues have been already identified in a phylogenetically distant gastropod Biomphalaria glabrata 55 , indicating that these receptors are conserved in this group. As a whole, these results allow us to assume that PPARγ would function as a heterodimerization partner of RXR in the penis of both B. globulosus and T. geversianus males, and that RXR is the one that triggers imposex onset. However, many aspects of the RXR/PPAR system still remain to be explored.
Our results must be taken into account when performing environmental monitoring studies using imposex as a biomarker, since the studies of morphological structures (i.e. penis or vas deference) could hide physiological responses to marine pollutants. In the case of T. geversianus, this is the only muricid species described that is less sensitive to TBT contamination, developing secondary sexual characters only at high TBT concentrations 41 .
RXR was previously reported to be expressed at the mRNA level in penis from the marine muricid gastropods T. clavigera 56 , N. lapillus 21 and P. pansa 30 , and at the protein level in T. clavigera 56,57 . However, to our knowledge, there is no information about PPARγ protein expression in marine gastropods. RXR and PPARγ interact in heterodimeric complexes after their activation by TBT 58 or 9cRA, the biologically active metabolite of vitamin A 59 . It has been hypothesized that signaling through both proteins has a key role in male and female seasonal reproductive development 29,60 and imposex, as previously mentioned. Whereas there is scant information regarding the physiological roles of RA in invertebrates, it is well known that signaling pathways through RA receptors exert a key role in embryo patterning and organogenesis in vertebrates 61 . 9cRA, is the natural ligand for mammalian RXRs, while the natural ligand of gastropod RXRs is currently unknown 62 . Vitamin A can either enter in a cascade producing retinal and RA, or undergo esterification to promote retinoid storage 63 . RA isomers were detected in testis, ovary and CNS of the caenogastropods N. lapillus and Nassarius reticulatus. However, they seem to be unable to store RA 27,64 . Finally, injection of TBT or 9cRA, into N. lapillus or H. trunculus males induces the outgrowth of reproductive structures 14,15 . All these data support the close relationship between retinoids and RXR/PPARγ in the development of male reproductive organs in marine gastropod mollusks.
The abnormal modulation of RXR signaling pathway by organotins seems to be the most accepted explanation of imposex development based on imposex induction assays together with the in vitro transcriptional activity of RXR 20,31 . Our findings, regarding the expression of RXR, contribute to the general understanding of the endocrine system in gastropods, and allow us to hypothesize that the degree of expression of RXR has direct implications on the sensitivity to imposex development.
We have also shown that imposex-affected female of B. globulosus collected from the polluted area (LPBH) of Nuevo Gulf, express RXR and PPARγ. These results are consistent with those obtained in species such as T. clavigera 56,65 , N. lapillus 66 , H. trunculus 15 , P. pansa 30 . However, our study is the first description of PPARγ protein expression in caenogastropod imposex-affected females. Regarding PPARγ actions, Pascoal et al. 67 showed that the PPARγ ligand Rosi elicited the same imposex response in N. lapillus as TBT. However, Giraud-Billoud et al. 68 reported that Rosi was not able to induce imposex in a phylogenetically distant gastropod Pomacea canaliculata. This difference indicates that different molecular mechanisms may regulate imposex induction in ampullariid and neogastropod species suggesting that the direct role of PPARγ in imposex deserves further investigation.
Males of B. globulosus that inhabit LPBH area are exposed to TBT 41,44 , PAHs 46 , trace metals and products from fishing industries 45,69 . These populations have increased oxidative stress responses compared to populations in CA beach 69 , indicating the negative effects of pollutants present at the harbour site on its physiological state. Our previous studies demonstrate the presence of high butyltins (TBT, DBT and MBT) levels in LPBH, both in sediments and in edible gastropods 41,44 . The sediments from LPBH in 2015 44 exceeded the TBT limit concentration established by international organizations, this probably continues up to the present time, as occurs in other countries of Latin America 10-12 . This may explain our findings of imposex-affected females in the area. It has been shown using in vitro assays 70 , that PAHs enhance the effect of natural ligands of retinoid signaling pathway, indicating that these environmental pollutants may influence the differentiation process and the embryonic development mediated by retinoids.
Differences in RXR expression between males from LPBH and CA could be related to the pollutants present in the harbor area. However, this should be tested in controlled experiments exposing normal individuals to TBT, PAHs or trace metals separately. The regulation of RXR expression by TBT seems not to be conclusive. Domínguez-Ojeda and colleagues, reported a down regulation of RXR induced by TBT in different tissues of P. pansa, while no changes were observed in penis of males 30 . Similar results were obtained in N. lapillus, H. trunculus and T. clavigera where no changes in RXR mRNA expression were observed using penis of males after TBT treatment 15,66,71 . The ideal correlation between mRNA-protein may be affected by highly dynamic phases, such as cellular differentiation or stress response 72 . In our case, despite not being able to directly associate RXR protein expression to any specific factor, we found a difference in RXR protein between both sites, but not for PPARγ, probably indicating a major role of RXR in the induction of the imposex phenomenon in these species and probably other around the world.
Overall, our results clearly show that differences in RXR male penises expression between B. globulosus and T. geversianus do not affect the normal development of secondary sexual organs. Future cloning and functional studies with B. globulosus and T. geversianus RXRs might reveal its real contribution to the observed differences in imposex development and penis formation between both species.