Anxiolytic activity of progesterone in progesterone receptor knockout mice

Abstract

Progesterone is an anxiolytic steroid that could play a role in the regulation of anxiety in women. However, the mechanism by which progesterone decreases anxiety is incompletely understood. Progesterone affects the function of the brain by two distinct mechanisms. Progesterone regulates reproductive behavior by activating intracellular progesterone receptors (PRs). In addition, progesterone is believed to influence neuronal activity through its conversion to allopregnanolone, a neurosteroid that acts as a positive allosteric modulator of GABA_A receptors. The extent to which the anxiolytic action of progesterone requires PRs is uncertain. In this study, we utilized PR knockout (PRKO) mice bearing a targeted null mutation of the PR gene that abrogates the function of both PR-A and PR-B subtypes to determine the requirement for PRs in the anxiolytic actions of progesterone. The absence of PR receptor protein expression in PRKO brain was confirmed by immunocytochemistry. In PRKO mice and their isogenic wild-type (WT) littermates, progesterone administration was associated with a dose-dependent rise in plasma allopregnanolone concentrations and corresponding anxiolytic effects in the elevated plus maze test. PRKO mice exhibited a greater anxiolytic response than WT animals although the allopregnanolone levels were similar in the two genotypes. Allopregnanolone also exhibited anxiolytic effects, but the magnitude of the response was similar in both genotypes. Pretreatment of PRKO mice with finasteride, a 5α-reductase inhibitor that blocks the conversion of progesterone to allopregnanolone, completely prevented the anxiolytic activity of progesterone, but had no effect on the response to allopregnanolone, demonstrating that allopregnanolone (or other 5α-reduced metabolites of progesterone) accounts for the response to the parent steroid hormone. These results provide direct evidence that the anxiolytic action of progesterone does not require PRs. However, PR activation by progesterone may influence the anxiolytic response since PRKO mice were more sensitive to progesterone.

Introduction

Progesterone has long been known to have anxiolytic activity in animals (Fernández-Guasti and Picazo, 1992, Freeman et al., 1993, Reddy and Kulkarni, 1996, Reddy and Kulkarni, 1997a, Reddy and Kulkarni, 1997b, Reddy and Kulkarni, 1997c, Friess et al., 1997, Frye et al., 2000, Frye et al., 2004) and man (Little et al., 1974, Kroboth and McAuley, 1997), although the underlying mechanisms have not been fully elucidated. Progesterone regulates diverse neuroendocrine functions by binding to progesterone receptors (PRs), intracellular receptor proteins that are members of the nuclear receptor superfamily of transcription factors (Etgen, 1984, Tsai and O'Malley, 1994, Mulac-Jericevic et al., 2000). Many of the behavioral effects of progesterone are believed to be mediated by PRs that are expressed in hypothalamic and extrahypothalamic cell populations within the brain. In particular, PRs in the hypothalamus and limbic areas are required for the induction of progesterone-dependent reproductive behaviors in females and are also involved in the regulation of maternal and aggressive behaviors (Parsons et al., 1982, Mani et al., 1997, Conneely et al., 2001, Levine et al., 2001). In addition, progesterone can influence behavior through its metabolite allopregnanolone, a neurosteroid that does not interact with PRs but which can rapidly influence brain excitability as a result of its activity as a potent positive allosteric modulator of GABA_A receptors (Rupprecht, 2003, Lambert et al., 2003). Studies in various animal models of anxiety have demonstrated that allopregnanolone has anxiolytic activity (Bitran et al., 1991, Wieland et al., 1991, Wieland et al., 1995, Carboni et al., 1996, Brot et al., 1997, Reddy and Kulkarni, 1996, Reddy and Kulkarni, 1997a, Reddy and Kulkarni, 1997b, Reddy and Kulkarni, 1999, Rodgers and Johnson, 1998, Frye et al., 2000, Finn et al., 2003). Moreover, several lines of evidence indicate that the anxiolytic effects of progesterone occur as a result of its conversion to allopregnanolone. Thus, progesterone-induced anxiolytic responses have been found to be correlated with blood and brain allopregnanolone levels and with the degree of facilitation of GABA_A receptor-mediated chloride current responses (Bitran et al., 1993, Bitran et al., 1995). In addition, anxiety measures are closely correlated with natural fluctuations in progesterone levels during the ovarian cycle and the corresponding changes in plasma and brain allopregnanolone (Schmidt et al., 1994, Wang et al., 1996, Rapkin et al., 1997, Rubinow et al., 1998, Genazzani et al., 1998, Frye and Bayon, 1998, Concas et al., 1998). Finally, treatment with a 5α-reductase inhibitor that prevents the conversion of progesterone to allopregnanolone eliminates the anxiolytic activity of progesterone (Bitran et al., 1995). However, the anxiolytic activity of progesterone is absent in some but not all anxiety models in mice deficient in type I 5α-reductase (Frye et al., 2004). Notably, there is no reduction in the anxiolytic effects of progesterone in the elevated plus maze, despite the fact that brain allopregnanolone levels are reduced by 60%.

Even if allopregnanolone is primarily responsible for the anxiolytic action of progesterone, PRs are activated when endogenous levels of the hormone rise or when progesterone is administered exogenously. The parent steroid could therefore play a role in the overall anxiolytic response by influencing gene expression relevant to the anxiety behaviors. In addition, although allopregnanolone does not bind directly to PRs (Gee et al., 1988, Rupprecht et al., 1993, Rupprecht et al., 1996), allopregnanolone is capable of indirectly influencing PRs as a result of its in situ conversion to 5α-pregnane steroids that do activate PRs (Rupprecht et al., 1993). Therefore, in the present study, we examined the possible contribution of PRs to the anxiolytic activity of progesterone using progesterone receptor knockout (PRKO) mice. PRs exist in two protein isoforms (PR-A and PR-B) that are transcribed from a single gene. In PRKO mice, both isoforms are functionally ablated by targeting the PR gene in the germ line (Lydon et al., 1995). PRKO mice showed no overt behavioral abnormalities compared with wild-type (WT) animals and the anxiolytic activity of progesterone was not diminished in the knockout animals, providing strong confirmation that PRs are not required for the anxiolytic actions of progesterone.