Elevation of brain allopregnanolone rather than 5-HT release by short term, low dose fluoxetine treatment prevents the estrous cycle-linked increase in stress sensitivity in female rats
Introduction
Premenstrual syndrome (PMS) and its more severe counterpart: premenstrual dysphoric disorder, blights the lives of millions of women worldwide (Steiner, 1997). In susceptible women the symptoms, which include angry outbursts, irritability and anxiety (ACOG, 2001), can be considered an exaggerated and inappropriate response to everyday acute stressful challenges. Surprisingly, given the enormity of the problem of PMS, there has been relatively little basic scientific study into its neurophysiological basis. What is clear from clinical studies is that PMS is dependent on cyclical variations in female sex hormones. Symptoms do not appear in anovulatory cycles (Backstrom et al., 2003). However ovulation itself is not the key factor since many women taking the combined contraceptive pill on a 21 day on, 7 day off dosing regimen, which prevents ovulation, also experience PMS-like symptoms, which peak during the 7 day drug free period (Kadian and O’Brien, 2012). In both cycling women and those taking the pill, symptoms occur at a time when blood levels of progesterone and estrogen, or their synthetic analogues, are in rapid decline.
In rats, withdrawal from long term dosing with exogenous progesterone at doses sufficient to raise progesterone to the high physiological range, precipitates increased anxiety (Smith et al., 1998, Smith et al., 2006). Spontaneously cycling rats in the late diestrus phase, when progesterone secretion declines sharply but estrogen secretion remains at a stable low level (Butcher et al., 1974), also become more fearful in an open field arena (Devall et al., 2009). Moreover, exposure of rats to 5 min of mild anxiogenic vibration stress (5 min vibration at 4 Hz whilst confined in a tube, Jørum, 1988) during late diestrus evokes a hyperalgesia, which is not seen when the animals are exposed to the same stressor at other stages of the cycle (Devall et al., 2009). These findings suggest that falling progesterone predisposes to an enhanced response to psychogenic stress. By analogy, the rapid decline in ovarian progesterone secretion in women during the late luteal phase might also provide a trigger for the enhanced responsiveness to psychogenic stress, which is a feature of the premenstrual period in many women (Nillni et al., 2011, Hoyer et al., 2013, Gollenberg et al., 2010). Thus whilst the cause of PMS in women is likely to be multifactorial, withdrawal from progesterone could be a significant contributory precipitating factor.
Progesterone passes readily through the blood brain barrier. The active agent that triggers the neuronal response to progesterone withdrawal in female rats is not however the native steroid hormone, but its neuroactive metabolite allopregnanolone (ALLO: 5α-pregnan-3α-ol-20-one or 3α,5α-tetrahydroprogesterone) (Gulinello et al., 2003, Smith et al., 1998). This progesterone metabolite ALLO is a potent positive allosteric modulator of the actions of GABA at GABAA receptors (Paul and Purdy, 1992) and its concentration in random cyclic female rat brain correlates with that of its precursor progesterone in plasma (Corpéchot et al., 1993). Ovarian secretion of both progesterone and ALLO decreases sharply at late diestrus (Ichikawa et al., 1974, Holzbauer, 1975), and so the concentration of ALLO in the brain will decrease in parallel. There is some endogenous production of progesterone and ALLO in the male rat brain (Cook et al., 2014). A similar situation appears to exist in the female rat, as evidenced by the persistence of these steroids in the brains of ovariectomised and adrenalectomised animals, but this is at concentrations around 9-fold and 40-fold lower, respectively, than those seen in the brains of intact random cycling rats (Corpéchot et al., 1993). Even in female rats at late diestrus, when ovarian progesterone secretion is low, brain concentrations of ALLO and progesterone were 10-fold higher than in males (Fry et al., 2014).
Whether naturally during the ovarian cycle or following administration of exogenous progesterone, withdrawal from ALLO, triggers upregulation of extrasynaptic GABAA receptors in the brain and consequent changes in excitability of neuronal circuits associated with anxiety (Gangisetty and Reddy, 2010, Griffiths and Lovick, 2005a, Griffiths and Lovick, 2005b, Gulinello et al., 2003, Lovick et al., 2005, Smith et al., 1998). The dynamic of the fall in brain concentration of progesterone appears critical because abrupt withdrawal from an exogenous progesterone-dosing regimen in rats precipitates an increase in responsiveness to anxiogenic stressors, whilst a gradually tapered withdrawal does not (Doornbos et al., 2009, Saavedra et al., 2006). In this respect it is interesting that in women, an association between clinical features of PMS and rate of decrease in progesterone during the luteal phase has been noted (Halbreich et al., 1986). Thus we reasoned that if the sharp fall in brain ALLO concentration that occurs at the end of the estrous cycle in rats is a trigger for the development of increased stress sensitivity, measures to produce a more gradual reduction in brain concentration of ALLO at the end of the estrous cycle should prevent the development of withdrawal-like symptoms.
In male rats and mice, the antidepressant fluoxetine (FLX) has been shown to induce an increase in the concentration of ALLO in the brain, which can be detected within 20 min of acute administration (Pinna et al., 2009, Serra et al., 2001, Uzunov et al., 1996). We therefore hypothesized that if FLX produces a similar effect in female rats in late diestrus, then short term dosing with FLX during this stage of the ovarian cycle should offset the rapid physiological fall in the concentration of ALLO. Further, that the stimulus for the development of withdrawal effects, which normally characterise this phase of the ovarian cycle, should be absent and changes in behavioral responsiveness to stress should not occur. To test this hypothesis we chose to focus on changes in neuronal responsiveness in the PAG, since we have already shown that estrous cycle-linked changes in GABAA receptor expression and neuronal excitability occur in this structure alongside changes in behavioral responsiveness (Lovick et al., 2005, Devall et al., 2009). Moreover electrical stimulation of the PAG evokes behavior that characterises responses to threatening and hence stressful situations in the rat and is sensitive to actions of anxiolytic drugs (Borelli et al., 2004).
Section snippets
Experimental procedures
The following is a brief outline of the methods used. Detailed descriptions are available as a Supplementary file.
Effect of fluoxetine on allopregnanolone concentration in late diestrus
Allopregnanolone was detected in the brain tissue of all animals. However, in the FLX-treated group (n=5) the concentration was more than double that measured in the saline-treated group (n=6) (⁎p<0.01, unpaired t-test, Fig. 1A).
Effect of fluoxetine on extracellular brain 5-HT concentration in late diestrus
Microdialysis samples for 5-HT measurement were taken after the second dose of FLX, i.e. at the equivalent time to when other rats were used for the behavioral testing or the collection of brain samples for ALLO measurement. One group of rats used for microdialysis
Discussion
In the present study, 5 min exposure to anxiogenic vibration stress induced hyperalgesia during late diestrus but not at other stages of the cycle, as shown previously (Devall et al., 2009). Interestingly, during nociceptive flexor reflex threshold testing in women (arguably also a mildly stressful procedure) a mild hyperalgesia has been demonstrated during the luteal phase, similar to our finding in rats (Tassorelli et al., 2002). In addition, in the present study the thresholds for fear-like
Role of funding source
Funding for this study was provided by MRC Grant G0700379 and CNPq grants 401898/2013 and CNPJ 63025530000104. JMS was in receipt of FAPESP Fellowship 08/11408-2. None of the funding agencies had a further role in study design; in the collection, analysis and interpretation of data; in the writing of the report or in the decision to submit the paper for publication.
Author contributors
The bulk of the experimental work was carried out by Adam Devall and Julia Santos who also analysed data. Jonathan Fry developed the protocol for steroid measurement with input from John Honour, carried out some of the experimental work and analysed all the data. Marcus Brandão advised on the work carried out in Brazil. Thelma Lovick conceived the study, contributed to experimental work and analysis of data and wrote the draft of the paper, with input from Jonathan Fry. All authors approved the
Conflict of interest
The authors declare no conflict of interest.
Acknowledgements
The work was supported by MRC Grant G0700379 and CNPq grants 401898/2013 and CNPJ 63025530000104. We thank Dr Milene Carvalho for advice on in vivo microdialysis. JMS was in receipt of FAPESP Fellowship 08/11408-2.
References (63)
- et al.
Descending facilitation from the brainstem determines behavioural and neuronal hypersensitivity following nerve injury and efficacy of pregabalin
Pain
(2008) - et al.
Effects of acute and chronic fluoxetine and diazepam on freezing behavior induced by electrical stimulation of dorsolateral and lateral columns of the periaqueductal gray matter
Pharmacol. Biochem. Behav.
(2004) - et al.
Neuronal excitability in the periaqueductal grey matter during the estrous cycle in female Wistar rats
Neuroscience
(2007) - et al.
Different patterns of freezing behavior organized in the periaqueductal gray of rats: association with different types of anxiety
Behav. Brain Res.
(2008) - et al.
Hyperalgesia in the setting of anxiety: sex differences and effects of the oestrous cycle in Wistar rats
Psychoneuroendocrinology
(2009) - et al.
Abrupt rather than gradual hormonal changes induce postpartum blues-like behavior in rats
Life Sci.
(2009) - et al.
Effect of an uptake inhibitor on serotonin metabolism in rat brain: studies with 3-(p-trifluoromethylphenoxy)-N-methyl-3-phenylpropylamine (Lilly 110140)
Life Sci.
(1974) - et al.
Neurosteroid withdrawal regulates GABA-A receptor alpha4-subunit expression and seizure susceptibility by activation of progesterone receptor-independent early growth response factor-3 pathway
Neuroscience
(2010) Descending modulation of pain
Neurosci. Biobehav. Rev.
(2004)- et al.
Co-localization of 5-HT 2A -receptor- and GABA-immunoreactivity in neurones in the periaqueductal grey matter of the rat
Neurosci. Lett.
(2002)