AR-beta was expressed in all tissues in both sexes and reproductive states. The ubiquitous AR-beta expression in adult stickleback tissues is largely consistent with a number of earlier studies in teleosts (e.g. [2, 5, 6, 9, 10].
The AR-beta mRNA levels in stickleback tissues range up to two orders of magnitude with the highest levels in the kidneys and the gonads. The kidney is a well-known androgen target tissue in sticklebacks. The male kidney hypertrophies in the breeding season and produces a glue-protein, spiggin, which is used in the building of the nest . Kidney hypertrophy/spiggin production is stimulated by androgens, especially 11-ketoandrogens, and is suppressed by castration [12–15]. The androgens act on the kidney level. Androgen treatments induce transformation of kidney secondary proximal tubule cells into glandular cells in vitro . Androgens induce spiggin synthesis in kidney cell  and tissue culture . Many differences in the structure of renal corpuscles in mature compared to immature males have been observed using electron microscopy . Kidney tissue cultures treated with T or 11KT also showed some of these effects, i.e. activated mesangial cells and podocytes . Androgens can stimulate the stickleback kidney also outside the breeding season  and in females, where otherwise kidney hypertrophy and spiggin levels are low [21, 22].
Similar to sticklebacks, the AR expression was higher in the kidneys than in many other studied tissues in zebrafish , male half-smooth tongue sole (Cynoglossus semilaevis)  and rice field eel (Monopterus albus) . For these and most fishes, no sex dimorphism in the kidney is known and the possible role of androgens has not been studied.
High AR mRNA levels in the testes have also been found in several previous studies on fishes (e.g. zebrafish (AR-beta) , Spinibarbus denticulatus (AR-beta) , sea bass (AR-beta) , three spot wrasse (Halichoeres trimaculatus) (AR-beta) ). This is not surprising since androgens are necessary for spermatogenesis in most studied vertebrates, including teleosts. A stimulating in vitro effect on spermatogenesis by androgens, preferentially by 11KT, has been shown in the Japanese eel [27, 28]. In contrast to most other studied teleosts, androgens inhibit spermatogenesis in stickleback [29, 30]. Males display secondary sexual characters and high levels of 11KT and T during breeding season when spermatogenesis is quiescent . Spermatogenesis commences at the end of the breeding season, when circulating androgens levels are very low . Administration of androgens, particularly 11-ketoandrostenedione (11-KA, which is converted extra-testicularily to 11KT ) at the end of breeding prevents the onset of spermatogenesis [29, 32].
In the stickleback, AR-beta mRNA levels were as high in the ovaries as in the testes. Similar findings were reported in the three spot wrasse . High mRNA levels of AR-alpha, but not of AR-beta were found in ovaries of rainbow trout . In zebrafish  and sea bass  there was higher level of AR-beta mRNA in the testes than in the ovary. Ovarian AR-beta mRNA levels were at most 1% of those found in testes in Spinibarbus denticulatus . It is not known what biological role(s) AR expression may have in the stickleback ovary or in female fish in general. Female teleost fishes, including the stickleback , display much lower levels than males of 11-KT, the most effective androgen in fishes . However, females have equal or even higher circulating levels of T than males in most studied teleosts , including breeding sticklebacks sampled in the field . Testosterone may have action(s) by itself or after conversion to more effective androgens. However, it may also be converted to estrogens. Indications for androgenic actions on the ovary are provided by effects of the anti-androgen flutamide on fathead minnows, Pimephales promelas . High doses of flutamide reduced the number of spawnings and eggs/spawning. Treated females had higher plasma levels of vitellogenin and T but not of 17β-estradiol, less oocyte maturation and more atretic follicles than control fish . The low maturation and high levels of vitellogenin may suggest an impaired uptake of vitellogenin into the eggs.
AR-beta mRNA levels were also high in the stickleback intestine. The intestine has been proposed to take over the kidneys' role as a freshwater osmoregulatory organ in mature stickleback males . The androgen-induced transformation of the kidneys into a glandular organ excreting spiggin also leads to a reduced ability to re-uptake ions and to excrete surplus of water . On the other hand a higher intestinal fluid secretion was found in mature than in immature stickleback males . Furthermore, the basal labyrinth, an intracellular membrane system characteristic for ion and water transporting epithelial cells, was more developed in the enterocytes in mature males and androgen treated females than in immature males and females . The high AR-beta expression in the intestine suggests that androgen effects are likely to be exerted directly on the intestine level. A role of the intestine in hyperosmoregulation is not known from other fishes. On the contrary, fish in seawater drink to compensate osmotic water losses and the water is absorbed by the intestine . In salmonids, sexual maturation and androgen treatments impair the hypoosmoregulatory ability , but it is not known whether this response involves the intestine. Expression of AR in the intestine has only been found in a few teleost fishes. AR expression was found in all investigated tissues, including the intestine in male half-smooth tongue sole .
AR-beta was present in the stickleback brain, but the levels were low. Androgens stimulate reproductive behavior in the male stickleback , but the effects may be exerted via a small proportion of the brain cells. AR(s) were also expressed in brains of several other fishes [2, 4, 5, 9, 25, 26, 40].
The AR-beta mRNA level in the stickleback liver was low. This is in general agreement with findings in the zebrafish . However, in Spinibarbus denticulatus , Mozambique tilapia  and rice field eel  liver AR mRNA levels were higher than in any other studied organs. Yolk proteins like vitellogenin are produced by the liver under stimulation of estradiol. High AR expression in the liver may suggest that also androgens are involved in the control of vitellogenesis. This is also supported by in vitro effects of androgen on the liver of Japanese eel . In this species, AR-alpha was found in the liver, whereas AR-beta was not .
In the stickleback, both pectoral and axial muscle contained low levels of AR-beta mRNA. AR(s) in musculature have also been found in some other fishes [6, 9, 10]. Male sticklebacks have larger pectoral muscle than females and use their pectoral fins vigorously when ventilating the nest . Castration reduced the weight of pectoral muscles but T or 11KT treatment did not restore it .
There was no significant effect of sex on AR-beta mRNA levels in any studied organ in the present study. There are very few other studies where AR mRNA has been studied quantitatively in both sexes of fishes. In zebrafish there were about twice as much AR-beta mRNA in testes as in ovaries and in male than in female muscle, whereas there was no significant effect of sex in brain, kidney, liver, skin and eye . AR-beta mRNA levels were measured in pituitary, brain, gonads and liver in male and female Spinibarbus denticulatus . The levels were clearly higher in testes than in ovaries, but there were lower levels in the pituitary of males than in females.
We found no significant differences in AR-beta mRNA levels in the kidneys between breeding male and female sticklebacks whereas in another study a 1.4-fold higher kidney AR-beta expression in mature males than in mature female sticklebacks was reported .
Although breeding and post-breeding fish, particularly in males, showed large differences in 11-KT and T plasma levels, there were no differences in AR-beta mRNA levels in most studied tissues, except for the kidneys. Unexpectedly, we found higher AR-beta mRNA levels in the kidneys of post-breeding compared to breeding males. Breeding males in the present study displayed well-developed hypertrophied kidneys and high T and 11KT levels. Conversely, none of the post-breeding males showed hypertrophied kidneys and their androgen levels were very low, indicating that they had left the reproductive state. High AR-beta mRNA levels in the kidneys of breeding males are likely necessary for the high spiggin synthesis. What function(s), if any, males may have of continuous high AR-beta expression in the kidneys shortly after the reproductive period is unknown.
These results suggest that androgens have a little, if any, effect on AR-beta expression in most organs and at least no positive effect in the kidney. No support for auto-regulation of AR-beta mRNA or protein levels (binding capacity) in stickleback kidneys was found when intact females and castrated males were treated with androgens i.e. T, 11KA or 5-alpha dihydrotestosterone . Furthermore, no effect of methyltestosterone on kidney AR-beta expression was found in intact female sticklebacks, although this treatment increased spiggin synthesis by five orders of magnitude .