Estrogen and exercise interact to regulate brain-derived neurotrophic factor mRNA and protein expression in the hippocampus Nicole C. Berchtold, 1,2 J. Patrick Kesslak, 1,3 Christian J. Pike, 1 Paul A. Adlard 1 and Carl W. Cotman 1,2,3 1 Institute for Brain Aging and Dementia, 1226 Gillespie Neuroscience Facility, University of California, Irvine, CA 92697±4540 2 Department of Neurobiology and Behaviour 3 Department of Neurology Keywords: ageing, Alzheimer's disease, BDNF, hippocampus, rat Abstract We investigated the possibility that estrogen and exercise interact in the hippocampus and regulate brain-derived neurotrophic factor (BDNF), a molecule increasingly recognized for its role in plasticity and neuron function. An important aspect of this study is to examine the effect of different time intervals between estrogen loss and estrogen replacement intervention. We demonstrate that in the intact female rat, physical activity increases hippocampal BDNF mRNA and protein levels. However, the exercise effect on BDNF up-regulation is reduced in the absence of estrogen, in a time-dependent manner. In addition, voluntary activity itself is stimulated by the presence of estrogen. In exercising animals, estrogen deprivation reduced voluntary activity levels, while estrogen replacement restored activity to normal levels. In sedentary animals, estrogen deprivation (ovariectomy) decreased baseline BDNF mRNA and protein, which were restored by estrogen replacement. Despite reduced activity levels in the ovariectomized condition, exercise increased BDNF mRNA levels in the hippocampus after short-term (3 weeks) estrogen deprivation. However, long-term estrogen-deprivation blunted the exercise effect. After 7 weeks of estrogen deprivation, exercise alone no longer affected either BDNF mRNA or protein levels. However, exercise in combination with long-term estrogen replacement increased BDNF protein above the effects of estrogen replacement alone. Interestingly, protein levels across all conditions correlated most closely with mRNA levels in the dentate gyrus, suggesting that expression of mRNA in this hippocampal region may be the major contributor to the hippocampal BDNF protein pool. The interaction of estrogen, physical activity and hippocampal BDNF is likely to be an important issue for maintenance of brain health, plasticity and general well- being, particularly in women. Introduction The role of estrogen in the ageing brain has become an important issue in health care. In women, estrogen levels decline abruptly at menopause which may contribute to the cognitive decline seen in normal ageing and Alzheimer's disease (AD) (Simpkins et al., 1997). Reduced estrogen compromises neuronal function, survival, and synaptogenesis (for review, see Wise et al., 2001), reinforcing the evidence that postmenopausal estrogen replacement can slow age-related cognitive decline and delay the onset of AD (for review, see Garcia-Segura et al., 2001). The time interval between estrogen loss and hormone replacement intervention may be important for biological activity but needs further investigation. In addition to potential health bene®ts from estrogen, there is a growing body of evidence indicating that exercise contributes to cognitive function (Kramer et al., 1999; van Praag et al., 1999) and is associated with reduced relative risk for developing AD (Friedland et al., 2001; Laurin et al., 2001). A common gene product regulated by both estrogen and exercise is brain-derived neurotrophic factor (BDNF), a key neurotrophin important in synaptic plasticity (for review, see McAllister et al., 1999) and neuroprotection (Lewin & Barde, 1996). BDNF mRNA is present in numerous brain structures with the highest expression found in the granule neurons of the dentate gyrus and pyramidal neurons in the CA3 and CA2 region of the hippocampus (with lower levels in the CA1) (Ernfors et al., 1990). In the hippocampus and cortex, brain regions critical in cognitive function, estrogen has direct and indirect modulatory actions on certain neurotransmitter systems (Kritzer & Kohama, 1999), regulates synaptogenesis and spine formation (Woolley & McEwen, 1994; McEwen et al., 1997), and is neuroprotective (Garcia-Segura et al., 2001). Some of the bene®cial effects of estrogen may be due to increased availability of BDNF: the BDNF gene contains an estrogen- response element (ERE) (Sohrabji et al., 1995), and BDNF gene expression is increased in vivo by estrogen (Singh et al., 1995). Similar to estrogen, BDNF can regulate neurotransmitter systems (MacRae et al., 1987; Fordyce & Farrar, 1991; Bucinskaite et al., 1996), modulate synaptic transmission, and enhance synaptogenesis (McAllister et al., 1999). In addition, the BDNF gene is sensitive to regulation by physical activity levels and learning (Neeper et al., 1995; Kesslak et al., 1998), behaviours that in turn are also in¯uenced by estrogen levels (McEwen & Alves, 1999). Thus, estrogen and exercise may converge on the regulation of BDNF. Correspondence: Nicole C. Berchtold, 1 Institute of Brain Aging, as above. E-mail: nberchto@uci.edu Received 17 August 2001, revised 2 November 2001, accepted 5 November 2001 European Journal of Neuroscience, Vol. 14, pp. 1992±2002, 2001 ã Federation of European Neuroscience Societies