Estrogen and the aging brain

As women advance in age, pregnancy and childbirth become increasingly dangerous and destructive. Perhaps to protect us, we women have evolved to be infertile later in life: our ovaries stop producing estrogen, causing our reproductive systems to gradually cease operations. Thus rendered barren, we can devote our maternal resources to mentoring and supporting our children and grandchildren. The rosy "grandmother hypothesis" is, however, not the only theory for the evolutionary origin of menopause.

The cessation of estrogen production also results in a number of debilitating symptoms, such as hot flashes, loss of short-term memory, and declining abilities to concentrate and learn new tasks, which would have put older women at greater risk for predation. Accordingly, some have hypothesized that menopause evolved as a way to "thin the herd," eliminating non-reproductive members of society and leaving food and other resources for the young. (Love you, Gumma!)

[This "culling agent" theory receives little support; the predominant theory as to why cognitive abilities decline is that conditions manifesting later in life (especially after reproductive age) are simply not subjected to the pressures of natural selection.]

Regardless of prehistorical reality, humans have evolved the propensity to thwart nature, creating the pharmaceutical industry and one of its many gifts: hormone replacement therapy (HRT). HRT does not rescue infertility, but is intended to mitigate the other lamentable effects of menopause, such as those impacting cognitive function.

The aging brain, while not suffering from notable cell death (except in conditions like Alzheimer's and Parkinson's Disease), is afflicted by significant changes in the connections (synapses) between neurons, within otherwise intact neural circuits. Certain molecules with essential roles in synaptic communication (e.g. glutamate receptors) change in quantity and location. These molecular changes are accompanied by significant structural alterations to the synapses themselves. Two regions display the greatest vulnerability to these changes: the prefrontal cortex (PFC), involved in attention and working memory, and the hippocampus, involved in many types of memory formation. Although these changes are inevitable concomitants of brain aging, they are exacerbated by the drop in estrogen levels experienced by women undergoing menopause, particularly in the PFC.

Estrogen, like all hormones, acts by traveling through the membrane of a cell to the nucleus, where it switches certain genes on or off, thereby regulating protein production. Of the many genes under the direct control of estrogen are the NMDA receptor (a key molecule for synaptic communication, in particular synaptic plasticity), elements of the cholinergic system (involved in attention and working memory), and genes that influence neuronal survival and structure. In particular, estrogen is known to enhance the number and strength of connections in the PFC of female rhesus monkeys which have had their ovaries removed ("ovariectomized," or OVX). The relevance to the human menopausal situation, however, involving both age and estrogen loss, was heretofore unknown.

A new study by John Morrison at Mt. Sinai School of Medicine investigated this issue by OVXing old and young rhesus monkeys, and treating half of each group with estrogen. The group then tested the monkeys on a task of short-term memory (STM), a component of working memory, in which the monkeys had to remember the location of an object after an increasing delay. They found that aged OVX monkeys which had not received estrogen treatment performed significantly worse than any of the other three groups (aged OVX + estrogen (E), young OVX + E, young OVX), indicative of significant cognitive decline. Moreover, the two groups of young animals performed equivalently, regardless of whether they received estrogen treatment, and the aged OVX + E group performed equally well as the former two. This surprising finding indicates that the estrogen treatment in the aged monkeys was sufficient to improve their cognitive function to levels comparable to their younger peers.

After cognitive testing, the researchers analyzed the brains of all monkeys, discovering that, in the PFC, estrogen increased synaptic density in both young and old OVX monkeys. Highest synaptic density was observed in young OVX + E monkeys, followed by comparable levels between young OVX and aged OVX + E, and lowest density in aged OVX without E. Moreover, estrogen treatment resulted in a significant increase in a particular subpopulation of synapses, which exhibit high dynamism and plasticity.

These findings indicate a complex interplay between estrogen and age, by which "young monkeys without [estrogen] can sustain excellent cognitive function against a background of dynamic spine plasticity." The one-two punch of age and estrogen loss, however, may be sufficiently destructive to impair an animal's cognitive function. By promoting the growth of new, dynamic synapses, estrogen may partially compensate for the effects of aging.

The implication with respect to HRT is that the timing of treatment is crucial. It may be important to begin treatment when ovarian hormone levels just begin to fall, at perimenopause, while synaptic plasticity mechanisms are still robust and resilient. Thus, this study contributes to the enormous body of HRT research (which currently consists of heaps of conflicting information). It has been suggested that the timing of hormonal intervention may underlie many of these contradictory data, and this study may lend some credence to this hypothesis and clear these cloudy waters.

Reference: Hao J et al. Interactive effects of age and estrogen on cognition and pyramidal neurosn in monkey prefrontal cortex. PNAS 2007 Jun 25 [Epub ahead of print].

Adult Entertainment

The biology of brain plasticity, including adult neurogenesis, synaptic plasticity, axon regrowth, and synaptic reorganization, is currently one of the most intensively studied areas of neuroscience. One of the burgeoning avenues of research for this field explores how altered plasticity may account for some of the behavioral and neural changes afflicting the aging brain, and is leading to efforts of fostering plasticity and thus "rejuvenate" the brain.

Perhaps the most popular products to emerge from this research are "brain games," which are activities designed to enhance cognitive function. These games have encountered a healthy bit of well-deserved skepticism from the scientific community; few of them have been validated by techniques even remotely "scientific" (by including, for example, controls), yet many make grand claims of improving some general notion of "intelligence" and well-being. There are, however, a few exceptions: "brain fitness" products emerging from within the scientific community, such as Posit Science, intended for people in their 60s and 70s, and Lumosity, which targets a younger population (i.e. baby boomers.) Since many functions, such as processing speed, working memory, and attention, begin declining around the age of 30, it seems reasonable to start on the early side.

Lumosity is a new program, so the games are still in "beta" phase and thus free, and these games are easily the most entertaining of any I've previously played (evoking behavior reminiscent of my childhood Tetris addiction.) More importantly, the games are inspired by research on human cognition; the company's head of neuroscience research studied with Jon Cohen at Princeton and John Gabrieli when the latter was at Stanford, among others, and there are a number of cognitive neuroscientists on the board of advisors. I recently met one of the founders at Stanford, and after discussing my research on adult hippocampal neurogenesis, I ended up joining these cognitive neuroscientists as a fellow scientific advisor. Anyways, the group at Lumos Labs performed a randomized, controlled study, which I can personally endorse, showing that Lumosity users improved on tasks of working memory and visual attention (there's an SFN poster and white paper available for your scrutiny as well).

Further scientific validation of the program's ability to improve various cognitive functions is certainly needed, and is in progress. Most importantly, of course, will be evidence that this sort of cognitive training can have long-term effects that translate to "real-world" functional improvements. In the meantime, the games are fun (with enticingly impressive high score lists) and certainly can't hurt.

Old, wise, and happy

As we grow older, we experience a number of cognitive changes, such as poorer working memory, declining ability to encode new memories, and slower processing speeds. By contrast, a number of critical abilities (short-term memory, autobiographical memory, semantic knowledge) remain stable. One of the major avenues of research for the cognitive neuroscience of aging explores how these behavioral changes correlate with changes in neural structure and function. Such studies, which rely heavily on neuroimaging techniques, have revealed that older adults have lower volumes of grey matter than do younger adults, primarily as a result of decreased synaptic density (i.e. the number of connections ("synapses") between neurons). Particularly affected are the prefrontal cortex (PFC), highly involved in processing speed, attention, and working memory, and medial temporal structures such as the hippocampus, which is involved in encoding information into episodic memories.

One ability which is not believed to decline with age is emotional processing. In fact, recent behavioral studies suggest that healthy older adults may actually perform better on tasks involving the processing of emotional stimuli than younger adults, and tend to have an enhanced experience of positive emotions and/or reduced experiences of negative emotions. To date, however, little is known of the structural and functional integrity of the neural regions associated with emotional processing, such as the anterior cingulate cortex (ACC, located in the middle of the brain right behind the PFC, may be important for one's conscious subjective emotional awareness), insula (important for "bodily" experiences of emotion, e.g. heart rate, breathing), and ventral striatum (involved in motivation and goal-directed positive emotion). How is the activation of these regions during emotional processing affected by age?

To understand the biology of such age-associated changes, Brian Knutson and Laura Carstensen at Stanford used fMRI to examine brain activity during emotional "incentive processing" tasks (i.e. anticipation of a loss or gain) in younger (19-27) and older (65-81) adults, and recently published their results online in Nature Neuroscience.

Participants viewed one of six cues, which displayed the amount of money that could be gained or lost on a certain trial (+$0, +$0.50, +$5.00, or - the same amounts). They were then presented with a target, and if they responded quickly enough they either gained or avoided losing the specified amount. Both age groups performed equivalently, earning similar amounts.

According to data reported by the participants, both younger and older adults felt similarly in anticipation of gaining money, but younger adults responded more strongly in response to anticipation of monetary loss. In other words, the older adults experiences less negative emotion in response to the same cues. This difference has been previously observed, but it may have been due to a bias in self-reports; thus, the researchers used fMRI to look for neural correlates of these behavioral differences.

They found that during reward anticipation (after participants saw the cue, but before they responded to the target), both younger and older adults showed equal levels of activation of the ventral striatum, anterior insula, and medial caudate (part of the dorsal striatum). During loss anticipation, however, younger adults had greater activation of the medial caudate and anterior insula than their elders. Thus, both affective and neural data indicate that older adults experience diminished negative emotions in response to loss anticipation, but retain their abilities to respond positively to reward anticipation. The asymmetry with which age affects the functioning of these regions is intriguing; clearly, they are still capable of normal levels of activation, but are somehow dampened during certain negative emotions.

The authors comment that the "age-related sparing of positive emotional experience may be related to efforts to optimize emotional experiences as one approaches the end of life. One aspect of this optimization may involve reducing negative arousal during anticipation of negative events." This sounds quite rosy to me, but the authors also mention that reduced emotional reactions to anticipated loss may have negative effects. Older people may have, for example, altered abilities of risk assessment, which may lead to sub-optimal decision making. Nevertheless, the overall enhancement of well-being may be worth the risk.

Reference: Larkin GRS, Gibbs SEB, Khanna K, Nielsen L, Carstensen LL, Knutson B. (2007). Anticipation of monetary gain but not loss in healthy older adults. Nature Neuroscience Apr 29; [Epub ahead of print]

Got nicotine?

As a pledge-signing member of my elementary school's rigorous D.A.R.E. program, I was regularly exposed to horrific photographs of blackened lungs and deteriorating hearts, and was utterly convinced that cigarettes ravaged and destroyed everything they touched. Thus when my dad started smoking when I was about 5, I (with my mom's encouragement) flushed pack after pack down the toilet until he quit. I still do not smoke, and I don't plan to, but I have grown a bit more discerning with respect to sensationalist demonization of various pleasure-enhancing drugs.

It turns out that of the 4,000 or so compounds in tobacco smoke, including a variety of carcinogens, and toxins such as carbon monoxide, heavy metals, and cyanide, at least one ingredient actually has some beneficial effects: nicotine. There is a large body of research showing that nicotine, the ingredient that drives people to addiction, improves cognitive function in humans and laboratory animals. The most robust effect demonstrated in human smokers is an enhanced ability to sustain attention to a task for a prolonged period of time, an ability inextricably linked to learning and memory. Of course, learning and memory involve a number of processes (acquisition, encoding, storage, and retrieval), but the ability to concentrate on particular stimuli and screen out the rest is critical for the success of this operation.

Nicotine's beneficial effects on these "higher" cognitive functions have prompted efforts to develop nicotinic treatments for diseases associated with cognitive impairment, such as Alzheimer's disease, Parkinson's disease, attention deficit/hyperactivity disorder, and schizophrenia. However, this area of drug development is impeded by the complexity of nicotine's actions, including the observation that cognitive improvements have only been reliably detected in either smokers or the cognitively impaired. In contrast, nicotine tends to have deleterious effects on cognitive performance in "normal" non-smokers. (Another factor hampering the development of nicotine-based therapies is that they offer pharmaceutical companies little potential for financial gain, as nicotine sources are easy to come by.)

So how does nicotine affect cognitive function? First, a bit about how neurons communicate with each other. Some people may be able to skip the next few paragraphs (which I've distinguished with altered formatting), but I've included them anyways because it's important to have at least a rough idea of this process in order to understand how nicotine works in our brains.

Neurons are functionally integrated in expansive neural networks, with each neuron receiving up to thousands of inputs from other neurons. However, the vast majority of neurons* are not actually physically connected to one another; there is a tiny gap that separates neurons, called a synapse.

When a neuron is activated, an electrical pulse (an action potential) travels down its membrane; the neuron is said to "fire" an action potential. When the action potential reaches the end of the neuron, it cannot traverse the synapse, but instead induces the release of chemicals which can. Once liberated from the "pre-synaptic" neuron, these chemicals (called neurotransmitters) navigate across the synapse and bind to specific receptors on the "post-synaptic" neuron. Once bound, the neurotransmitters induce one of many physiological changes: they can make it easier to fire an action potential ("excitatory" neurotrasmitters), more difficult to fire an action potential ("inhibitory" neurotransmitters), or modulate the firing rate or other behavioral properties of the cell.

An overwhelming number of pre-synaptic neurons, all of which are sources of neurotransmitters, impinge on a single post-synaptic neuron, yet the latter responds with a binary decision: fire or don't fire. The cell creates order from this chemical deluge by performing a complex, time-dependent summation of all of its inputs; if it receives a sufficient number of excitatory inputs within a reasonable time window, it will fire an action potential and release its own neurotransmitter, passing the information along the circuit.

Each neurotransmitter can bind to a number of complementary receptors. One of the receptors for a neurotransmitter called acetylcholine (ACh) happens to also bind and respond to nicotine, which is not naturally present in the body. Thus when a post-synaptic neuron containing these particular receptors (called nicotinic ACh receptors, or nAChRs) is exposed to nicotine (as in when someone smokes a cigarette), it behaves as if it has been influenced by ACh; i.e. to an individual nAChR, nicotine and ACh are indistinguishable.

However, there is a crucial difference at the circuit level: ACh is regulated by your body, so it is typically released in small amounts by specific subsets of neurons at any given time. In contrast, nicotine, entering your body from an external source, can potentially act at all nAChR-bearing neurons simultaneously, leading to widespread activation and an assortment of consequences (including the release of other neurotransmitters, such as dopamine, endorphins, and ACh itself).

Turns out that the prefrontal cortex (PFC), a brain structure with a critical role in learning and memory, contains an abundance of nAChRs. This area receives information from all of the senses, and aids the learning process by directing attention to a limited set of input streams at a time. Like many cortical synapses, the excitatory synapses in the PFC are plastic, capable of undergoing systematic changes in synaptic strength/efficacy.

These changes are thought to underlie the processing and storage of information in neural circuits, and for that reason take place in a functionally relevant manner (i.e. one that is dependent on the activity of that particular synapse). Specifically, the robustness and direction (stronger or weaker) of the change in synaptic strength is dependent on the precise timing of pre-synaptic inputs and post-synaptic action potentials (also called "spikes"). This temporal correlation gives the process its name: spike-timing-dependent plasticity (STDP). According to the rules of STDP, a synapse with a high temporal correlation between pre-and post-synaptic activity will strengthen.

Importantly, excitatory synapses in the PFC change during working-memory related tasks, implicating the PFC in these cognitive behaviors. It is likely that nicotine's effects on attention and working memory are effectuated at the nAChR-containing synapses of the PFC, but the mechanistic changes are unknown. Moreover, it is unclear how these synaptic changes affect the functional properties of the circuit underlying these cognitive processes.

A group from Amsterdam, led by Huibert Mansvelder, published a study that explored the cellular and synaptic mechanisms of nicotine's actions in the most recent issue of Neuron, with a focus on how nicotine affects STDP in the PFC.

The scientists cut rat PFCs into slices, and induced STDP by electrically stimulating pre- and corresponding post-synaptic cells simultaneously. After repeating this paired stimulation (50 times), the synapse becomes "potentiated," meaning the pre-synaptic neuron becomes more effective at stimulating the same post-synaptic cell than it was before the procedure; i.e. the synapse is "stronger".

When nicotine was applied to the solution bathing the slice, this potentiation failed to occur. The blockade of STDP could be overcome, however, by increasing the electrical activity of the post-synaptic cell, indicating that the pairing procedure was less effective, but not defunct. The group found that nicotine's net effect was to enhance the release of a major inhibitory neurotransmitter, called GABA. In the context of the synapse, the post-synaptic neuron would thus be bombarded by copious amounts of GABA, which would then dominate the input summation. These actions decrease the likelihood that the post-synaptic neuron will fire, thereby interfering with the concomitant activation of both synaptic partners and interfering with STDP.

So how does impeding STDP, a process thought to provide the cellular foundation for an adaptive nervous system, enhance cognitive function in certain individuals? This question brings us back to attention--the ability to concentrate on relevant stimuli while ignoring that which is irrelevant. During PFC-based cognitive behaviors, the neural activity in the PFC may increase to distracting levels. By enhancing inhibitory neurotransmission, nicotine may enhance the "signal-to-noise" ratio, thereby improving attention selectivity. This may be particularly beneficial to smokers who are accustomed to high nicotinic stimulation, as well as individuals with cognitive impairment, as both these groups are functioning suboptimally in the absence of nicotine.

And what about normal individuals, whose cognitive functioning is often impaired with nicotine? Such "drug-free" individuals are probably already performing at or near their optimal level of performance. As a result, increasing nicotinic stimulation and interfering with STDP will have negative effects in most situations. It is possible, however, that even in "normal" individuals, nicotine may enhance cognitive function under extreme task demands. During such tasks, which necessitate intensified attention over a prolonged period of time, optimal performance may be facilitated by nicotinic stimulation.



*My explanation of synaptic transmission describes, specifically, a chemical synapse. These are the predominant form of synapses in the brain, but neurons can also be connected by channels called "gap junctions." The openings of these channels allow ions to flow from one neuron to the next, enabling electrical signals to pass directly between neurons. This type of connection is called an electrical synapse, and the transmission of information is much more rapid than at chemical synapses.