Environmental estrogens impact primate brain

In the early 1990s, a series of papers from Bruce McEwen's laboratory demonstrating that estrogen (i.e., estradiol) increased spine and axospinous synapse number in hippocampus profoundly altered our view of both synaptic plasticity and estrogen's role in nonreproductive brain functions such as learning and memory (1). “Axospinous synapses” are excitatory synapses on the spines of cortical pyramidal cells mediated by glutamate. These synapses represent the major site of excitatory transmission in cortical circuits and are modified in the context of learning and memory (2).

Thus, it is not surprising that estradiol enhances spatial memory, a hippocampus-dependent task, in female rats (3), although these cognitive effects, as well as the synaptic enhancement by estradiol, are attenuated in aged rats (4, 5). In recent years, such studies have been extended to nonhuman primates, demonstrating that estradiol increases spine and axospinous synapse numbers in monkey hippocampus (6, 7) and dorsolateral prefrontal cortex (8)—the neocortical region responsible for the highest level of cognitive function in monkeys and humans.

Multiple estradiol treatments improve cognitive performance in ovariectomized female monkeys (9, 10), and, unlike rats, aged monkeys remain responsive to estradiol with respect to both synaptic enhancement and cognitive performance (8, 9). These preclinical data have taken on increased importance in light of controversies in the clinical literature regarding menopause, estrogen treatment, and cognitive aging. However, more synergy currently exists between the animal and human data regarding estrogen and cognitive aging than appeared to be the case a few years ago (11, 12). This issue of PNAS contains an article by Leranth et al. (13) that is highly relevant to an equally important, yet controversial, clinical arena regarding estrogen's effects on the brain: the potential for estrogenic compounds from the environment to disrupt estrogen's natural role in regulating cortical structure and function.

Bisphenol A Blocks Estrogen-Induced Synaptogenesis

The report by Leranth et al. (13) focuses on bisphenol A (BPA), a common xenoestrogen (i.e., a manufactured estrogenic compound not generated in the body) that is present in some plastics. There appears to be little doubt that humans are exposed to BPA on a daily basis, although the health consequences of such exposure remain controversial. The authors demonstrate in a nonhuman primate model that a daily dose of BPA considered within the safe range (50 μg/kg) completely blocks the estradiol-induced increase in axospinous synapses in three distinct fields of hippocampus (CA1, dentate gyrus, and CA3), as well as in prefrontal cortex, in ovariectomized young adult female African green monkeys (13). Although the authors did not provide behavioral data, the fact that all four regions are affected suggests that the potential cognitive effects could be far-reaching, notwithstanding that young-adult monkeys are fairly resistant to cognitive decline resulting from ovariectomy, even with significant spine loss (8). As reviewed in Leranth et al., numerous behavioral studies in rodents have revealed abnormalities stemming from perinatal exposure to BPA. In addition, the same team has shown that BPA can block the estradiol-induced synaptic increase in ovariectomized female rats (14). However, the present study is the first to demonstrate this effect of BPA in a nonhuman primate model, which is of particular relevance to issues surrounding human exposure to BPA. In addition, in this study the monkeys were young adults, demonstrating that the capacity for BPA to disrupt estradiol's effects on neural circuitry goes beyond the vulnerable developmental stages.

Humans are exposed to BPA on a daily basis.

Potential Impact on Synaptic Function

A disruption of axospinous synapses of the magnitude reported by Leranth et al. (13) would likely have a profound effect on excitatory (glutamatergic) circuits in both hippocampus and prefrontal cortex. Spines and their axospinous synapses are highly plastic at both the molecular and structural levels (2). They contain multiple families of glutamate receptors (e.g., AMPA and NMDA) that interact in a complex fashion across receptor families, and each subtype interacts with hundreds of proteins that provide the receptors with numerous potential signaling cascades. The NMDA receptor, in particular, is known to play a critically important role in memory mediated by hippocampus. In fact, the effects of estradiol on synaptic morphology are NMDA-receptor-dependent (1), although likely mediated through both nuclear and local activation of estrogen receptors ERα and ERβ, both of which are present in the target synapses (15, 16). In addition, estradiol affects multiple signaling pathways in hippocampal synapses that have long-term functional consequences, such as the Src/ERK/CREB signaling cascade (17), the P13K/Akt pathway (18), and several other signaling cascades linked to synaptic plasticity and learning and memory (19). Spine plasticity and spine turnover in adults are very dynamic processes and likely are essential components of the synaptic events necessary to learn new tasks and form new memories. It has been demonstrated that ovariectomized rhesus monkeys treated with estradiol have both increased spine number in CA1 and prefrontal cortex and enhanced cognitive function on tasks mediated by these regions (8, 9). In addition, it has been proposed (i) that the estradiol-induced cognitive enhancement is a reflection of increased synaptic plasticity rather than simply the number of synapses (8) and (ii) that the spine and synapse plasticity is linked to several signaling cascades activated by estradiol. Thus, disruption of the synaptic effects of estradiol by BPA would likely have significant effects on the synaptic health and signaling dynamics of both hippocampus and prefrontal cortex.

Key Unresolved Issues

The effect of BPA on estradiol-induced synaptogenesis in these regions is clear and compelling; however, the repercussions for cognitive performance remain to be determined. Although one might assume that such powerful synaptic effects in hippocampus and prefrontal cortex would have detrimental effects on cognition, such deficits will likely depend on the age of the monkeys when ovariectomized and treated. Young monkeys display the same estradiol-induced spine/synapse increases as aged monkeys, but only the aged monkeys display robust cognitive deficits in the absence of estradiol (8). This is likely because aged monkeys experience the “double hit” of synaptic effects that result from aging and those that result from estradiol depletion, whereas young monkeys have sufficient synaptic resilience to perform well in the absence of circulating estradiol.

The duration of the BPA effect is also unclear and again is likely to be highly dependent on the age of the monkeys. Numerous studies have reported long-term behavioral effects of perinatal exposure to BPA in rodents (13), and such effects are presumably due to BPA's disruption of the appropriate development of neural circuits. It would be of great interest to extend such developmental analyses to the nonhuman primate model. The observations reported by Leranth et al. (13) are from ovariectomized young adult females treated with estradiol only, or with estradiol plus daily BPA, for 28 days after ovariectomy and euthanized at the end of the 28-day treatment period. We do not know how long the BPA effect would have endured if the animals receiving estradiol and BPA had been switched to estradiol only for a period before euthanasia. Such data would be valuable because it is important to know whether the synapses remain plastic and responsive to estradiol if exposure to BPA is decreased or eliminated. In addition, the potential for recovery if BPA exposure is discontinued may also depend on age because the affected synapses are also vulnerable to aging and perhaps are less able to recover in an aged animal. Such studies of BPA across the lifespan are essential follow-up analyses in the nonhuman primate model because they will provide valuable insight into the potential danger of exposure in humans at different ages and the potential ability of humans to recover if exposure is decreased.

As discussed by Leranth et al. (13), the mechanism by which BPA inhibits estradiol-induced synaptogenesis is unclear. Estradiol has numerous paths through which it can affect synaptic attributes, with at least two receptors that function both at the synapse and in the nucleus, and time frames for responses to estradiol that range from minutes to days. As pointed out by the authors, BPA has low affinity for both ERα and ERβ and thus might exert its effects through one of the many signaling pathways that estradiol activates (17–19). In addition, synaptogenesis and the function of axospinous synapses in hippocampus and prefrontal cortex are modulated by monoamine and cholinergic afferents to these regions. These systems are also responsive to estradiol treatment in monkeys (20) and may be directly affected by BPA.

Importance of the Nonhuman Primate Model

The article by Leranth et al. (13) points to many opportunities to investigate further the potential for xenoestrogens to disrupt the modulation by estradiol of cortical synapses and cognitive processes in a nonhuman primate model. This is an issue of great clinical relevance, and the nonhuman primate is well suited for these investigations for several reasons. First, the present study and others (8, 9) have implicated the prefrontal cortex as a major target of estradiol, and the nonhuman primate is an excellent model of human cognitive processes mediated by the prefrontal cortex. Second, endocrine physiology and neuroendocrine processes are very similar in monkeys and humans, which may be particularly relevant as the effects of BPA are tested at different life phases (perinatal, puberty, young adult, and aged). Third, it is often difficult to relate directly the neurological effects of a given exposure level of an environmental toxin in rodents to levels that humans are likely to experience. The fact that these results were obtained in nonhuman primates exposed to levels of BPA considered in the safe range for humans is cause for concern and warrants further analysis in both humans and nonhuman primates.