In teleost fish, the control of the spermiation process (hydration of semen and release by sperm duct) and the maturation of spermatozoa is only partly understood. The present study explored the potential involvement of 11-deoxycorticosterone (DOC), a mineralocorticoid, on the male reproductive tract function at the time of hydration and excretion of milt, in rainbow trout Oncorhynchus mykiss, a cyclic teleost fish.
We show that DOC plasma levels are particularly high around the time of spawning in mature male O. mykiss (Figure 1). DOC plasma levels peak around November in an autumn spawning strain and around May in a spring spawning strain, demonstrating that these changes are not "seasonal" variations. To our knowledge, this is the first report of plasma DOC up-regulation (10–50 fold) during a transitory period at the end of the male reproductive cycle in any teleost species. DOC has previously been reported in the blood plasma of one mature male O. mykiss and in mature winter flounder Pseudopleuronectes americanus, but in these studies DOC was not monitored in immature fish or outside the spawning period [27, 35]. Interestingly, this steroid was found to increase 38 fold in the blood of Tilapia aurea females after the initiation of spawning to reach values up to 22 ng/ml . By contrast, DOC was not detected in ovulated common carp Cyprinus carpio  or in pre-spawning Pacific hagfish Eptatretus stouti and sea lamprey Petromyzon marinus , and no difference in DOC plasma concentration was detected between gravid and sexually regressed female catfish Heteropneustes fossilis . These inter-species differences require further investigation. Nevertheless, the progressive increase in DOC observed here at the beginning of the spawning season means that we cannot rule out a direct or indirect role of DOC during trout spermiation.
In this study, plasma 17alpha, 20beta-dihydroxyprogesterone (MIS) also rose dramatically at the onset of spermiation and in close relation to the spermiation index (Figure 1). This confirms previous data [6, 19]. In the mature male population, peak values of the average DOC blood plasma levels were observed when all mature males had just started to excrete milt, and just before the peak values of MIS. The elevation of MIS and DOC coinciding with milt release is in agreement with the hypothesis that both steroids could be involved in sperm production or maturation in trout.
In mammals, the mineralocorticoid receptor (MR) main function is to regulate ionic homeostasis . The O. mykiss MR mRNA was detected in the spermiduct and in the testis at all studied reproductive stages (Figure 1). The relative abundance of this transcript was lower in mature testis compared to immature ones, probably due to a dilution effect linked to a decrease in the somatic cell/germ cell ratio during maturation. Such dilution effect has previously been observed for proteins or transcripts expressed in the testicular somatic cells [28, 40, 41]. Indeed, immunohistochemistry localized the rtMR to cells at the periphery of the seminiferous tubules (Sertoli and/or peritubular cells), but germ cells did not express the receptor (Figure 2).
At the end of the cycle, we further demonstrated an increase of testis rtMR mRNA abundance, particularly in mature males at the beginning of spermiation, just after the rise in plasma DOC (Figure 1). This observation is interesting since DOC strongly activates rtMR in vitro  and therefore is a potential ligand of the rtMR. The apparent relation between DOC plasma level and rtMR mRNA abundance is the first observation suggesting a physiological link between DOC and the rtMR. But, proving that DOC is the rtMR physiological ligand, the low plasma DOC level measured in immature fish would raise the question whether DOC is able to activate the rtMR in this physiological state. In that way, the higher rtMR mRNA observed in immature fish might increase the gonadal receptivity for DOC. Further studies are needed to understand DOC action outside the reproduction period.
An expression of MR has been described in mammalian testis, and the receptor has been localized to Leydig and Sertoli cells . This study is the first to report the cellular localization of MR in the reproductive tract of a fish species. Our immunolocalization of rtMR shows that mineralocorticoids could act on Sertoli cells and/or peritubular cells, which would be consistent with the involvement of these cells in water and ionic exchange with the germ cell compartment. In addition, presence of rtMR along vas deferens epithelium further argues for its implication in sperm hydration and ionic composition. Indeed, during fish spermiation, the sperm excretion is accompanied with important aqueous and ionic exchanges in the testis and the vas deferens [43, 44]. Variation of rtMR expression and localization during spermiation suggest that the rtMR might be involved in the endocrine control of this process.
In this study, the elevation of DOC and MIS coinciding with milt release is in agreement with the hypothesis that both steroids are implicated in trout spermiation. We tested the influence of DOC and MIS in vivo supplementation on several sperm parameters. In the present study, the initiation of spermiation was shown to be significantly advanced by MIS (Figure 3). In agreement with our observation, injections of MIS also induced precocious spermiation in amago salmon and brook trout [11–15]. The rather limited effect of MIS in our study might be explained by the relatively small increase of blood plasma levels of MIS obtained with the implants (from 0.7 ng in controls to 5 ng/ml, compared to the blood concentrations of 40 – 50 ng/ml observed in vivo during spermiation). In our study, DOC supplementation alone was not effective to trigger initiation of milt excretion, and did not appear to act synergistically with MIS on this parameter. This observation leads us to think that DOC itself is not strongly involved in the initial induction of milt release.
In previous studies, treatments with GnRH or MIS have induced an increase in milt volume in different fish species [2, 3, 10, 16, 45]. In our hands, neither DOC nor MIS significantly increased sperm volume in O. mykiss. We cannot exclude that the lack of effect is related to the specific experimental conditions of this study, but MIS injections have been found ineffective to increase milt volumes also in spermiating males of other salmonids [18, 19].
While Baynes and Scott  have found some positive correlations between blood plasma MIS levels and seminal fluid sodium/potassium concentrations in O. mykiss, MIS administration did not stimulate ion transport in the vas deferens in a previous study with brook trout . In our experiment, no effect of DOC and MIS treatments was observed on osmolality or sodium/potassium concentrations in the seminal fluid, and the correlations reported by Baynes and Scott  remain to be explained. Finally, in some teleosts it has been suggested that progestin receptors exist in spermatozoa, and a direct effect of progestins on a sperm carbonic anhydrase and in the increase of pH in seminal plasma has been proposed that could be involved in sperm maturation [46, 47]. However, results obtained in our study do not support a strong effect of DOC or MIS on the seminal fluid pH and spermatozoa motility.
Despite the absence of a significant effect on milt volume, DOC and MIS together significantly reduced the spermatocrit, whereas individually they did not (Figure 4). In vivo treatments with MIS in sea plaice Pleuronectes platessa, Atlantic halibut and Japanese eel Anguilla japonica support the role of MIS in the hydration of the milt [14, 16, 17]. In our in vivo study, MIS supplementation alone was not effective. This may have been due to the fact that only a moderate increase of this steroid in blood plasma was induced. However, our results suggest an interaction between DOC and MIS, resulting in increased milt hydration. To test whether this could possibly involve an indirect effect of DOC through a modulation of steroidogenesis, we investigated this possibility by studying MIS production in a tissue explant system. Cortisol, known to inhibit androgen production (see below), was also tested. It revealed that, in vitro, high concentration of DOC inhibited basal and LH stimulated MIS production (Figure 5). Interestingly, cortisol, within a physiological range of concentrations, also strongly reduced MIS production. Exposure to stress or cortisol has been shown previously to disrupt reproductive processes in fish, in particular by depressing gonadal steroid hormone levels [48–51]. To our knowledge, this is the first demonstration of cortisol and DOC effects on MIS production. It should be noted that DOC is a precursor of cortisol in fish; thus it could influence MIS production directly, or act after being metabolized into cortisol. With respect to the latter possibility, it is worth noting that 11beta-hydroxylase, which is the final enzyme of cortisol biosynthesis, shows high expression levels in testis . On the one hand, our observations refer that corticosteroids and MIS signalling pathways could interact during gonad final maturation . On the other hand, however, if MIS has a stimulatory role in spermiation, the down-regulation of MIS by DOC does not readily explain a positive role of DOC in this process. An alternative explanation, based on the immunolocalization of rtMR in this study and our previous finding that DOC is a strong agonist of the rtMR in vitro  is that DOC possibly acts via the rtMR directly on the seminiferous tubule and the efferent duct epithelium, affecting mechanisms related to water and ion exchange. To further elucidate the mechanism of DOC actions in the male teleost gonad, it would be interesting to investigate the potential involvement of the enzyme 11beta hydroxysteroid dehydrogenase 2, which metabolizes cortisol into cortisone, thus potentially allowing DOC to access the MR despite the greater abundance of cortisol in plasma. Finally, we speculate that DOC's effects during spermiation could also be mediated through a pathway involving a nuclear receptor similar to the one that was characterized in seatrout ovaries and showed high specificity for C21 progestagens and 11-deoxycorticosteroids ; however, such a receptor has not yet been characterised in the O. mykiss testis.