They may have many causes and manifestations, but diseases such as Alzheimer’s, Parkinson’s, and stroke all end the same way: with the death of essential brain cells. We urgently wish to prevent the death of cells, but it can also be natural and necessary.
Neuroanatomist James Hutchins looks at “apoptosis,” a fundamental process by which our cells are actually programmed to die, an essential step in early development and perhaps also throughout life. If genetics research can ﬁnd the switch that turns apoptosis off, our concept of aging may undergo a stunning change. But before we rush to ﬂip that switch, warns Hutchins, we have some knotty problems to ponder.
Few things are more exquisitely frustrating than losing control of something precious. A moment’s inattention and the priceless wine glass slips from our grasp. We watch, horriﬁed, as it seems to descend in slow motion.
For families, caregivers, and doctors watching the progress of a neurological illness, the same feelings are evoked. As a brain deteriorates from Alzheimer’s disease or Parkinson’s disease or stroke, the victim’s unique personality seeps away. Like Jimmy Stewart at the beginning of Hitchcock’s Vertigo, we watch our loved ones slip from our grasp; like Stewart, we wonder, in agony, “Isn’t there something more that can be done?”
Each year, neurological diseases claim in the neighborhood of half a billion victims worldwide. In these diseases, neurons— the brain cells responsible for receiving, processing, and transmitting information— die and are not replaced.
The current scientiﬁc literature makes one thing plain. As yet, there is no unifying hypothesis for the causes of Alzheimer’s disease, Parkinson’s disease, or many rarer degenerative brain diseases that cause dementia. Is each only one disease—that is, attributable to a single cause—or many? My hunch is that they are more like cancer: the outcome not of a single disease agent or genetic malformation but of many steps along the path to disease.
Recently, however, we have begun to understand that a process known as programmed cell death, or apoptosis, may be the ﬁnal common pathway in many, if not most, of these neurological diseases. In Alzheimer’s, some initial imbalance or insult, or damage accumulating over years, ﬁnally leads to death by apoptosis of neurons in the brain centers responsible for learning, memory, and cognition. In Parkinson’s, neurons in the brain’s substantia nigra (“black substance”) are lost, probably through apoptosis. These neurons normally release dopamine, which initiates, controls, and smooths body movements. With the loss of dopamine comes an increase in the characteristic tremor, loss of mobility, and disorganized thought (dementia). In ischemic stroke, resulting from a loss of blood supply, a relatively small number of neurons at the point of greatest damage die. While this central core undergoes an uncontrolled cell death, the much larger area known as the penumbra (“shadow”) is destroyed by apoptosis, rather than by direct loss of glucose and oxygen.
Consider two broad strategies for treating these diseases. In one strategy, it is as though we are trying to clean up the many piles of ﬂammable brush before they are ignited, setting off a single, unmanageable forest ﬁre. In the other, we are attempting to prevent the piles from kindling and then uniting into a single, unmanageable ﬁre. In this second approach, we must ﬁnd the ﬁnal common pathway, the ﬂame that is the end result of each of many separate biological events. For many neurodegenerative diseases, that ﬁnal common pathway is apoptosis.
If, by stopping apoptosis, we can halt the progression of the disease, why not do so? Controlling apoptosis may offer real hope, I believe, but we should not rush to eliminate a process completely that we are just beginning to understand. There are hints, even now, that apoptosis may serve a critical purpose throughout life, not just during development, as originally thought.
WHAT IS APOPTOSIS?
Apoptosis was ﬁrst fully described as recently as 1972 by John Kerr, Andrew Wyllie, and Alastair Currie.1 Its name (usually pronounced “ay-poe-toe-sis,” though sometimes as “ay-pop-toe-sis”) comes from the Greek word for “a falling off” and is meant to evoke the dropping of leaves in autumn: a natural event in the passing of the seasons. Often, the alternative term “programmed cell death” is used, perhaps because it is more evocative of the process, but more likely because it is easier to pronounce.
With apoptosis, cells do not die randomly, and they are not killed. They commit suicide; they “choose” to die.
Although initially discovered in the immune system, apoptosis has now been studied in other locations, including the nervous system. Once considered arcane and rare, apoptosis is now known to be a ubiquitous feature of both the developing and adult organism. For example, the human hand has ﬁngers because of apoptosis. In the embryo, ﬁngers do not grow as extensions of the palm. A primitive, paddlelike structure forms ﬁrst, then cells occupying a location corresponding to the spaces between ﬁngers begin to die; as they are removed or fall away, the familiar structure develops. Clearly, if we want ﬁngers so that we can manipulate our environment, we want this cell death to occur. It is “appropriate.”
A study by Ron Oppenheim and his colleagues gives us an example of apoptosis in the early development of the nervous system.2 One portion of the spinal cord, the ventral horn, contains about one and a half as many neurons as the adult will need. Normally, these neurons die and are removed early in development. Their death seems to depend on an organized pattern of activity; if communication between nerve and muscle is blocked by the drug curare (also used by South American indigenous people to poison arrow tips), the “extra” neurons persist. Again, this is a good example of appropriate cell death, and a process that blocks it is abnormal.
With apoptosis, cells do not die randomly, and they are not killed. They commit suicide; they “choose” to die. Stated another way, apoptosis requires energy and the activation of speciﬁc genes in a particular sequence.
A good contrast to apoptosis is necrosis —cell death due to injury or severe metabolic disruption. Cells of the developing hand die so that ﬁngers may form (apoptosis); if a ﬁnger is severed, the cells die but no greater good to the organism results (necrosis). Neuroscientist Steve Estus has introduced the useful terminology “appropriate” versus “inappropriate” cell death.3 Apoptosis is appropriate in the formation of a hand or nervous system, but inappropriate if it causes the death of cells that the organism needs for proper function.
THE SUICIDE OF A CELL
What do we know about the complex operation of apoptosis in a cell? Several details have yet to be worked out, but the ﬁrst event in apoptosis seems likely to be a phenomenon in nerve cells that is called the “mitochondrial permeability transition.”
Neurons consume huge amounts of energy. The brain, at about 2 to 3 percent of body weight, consumes about 20 to 30 percent of the glucose and oxygen carried by the blood. The energy “currency” adenosine phosphate (ATP) is always in short supply in neurons. It is estimated that neurons turn over their supply of ATP every two minutes. Put another way, neurons have only two minutes’ salary in reserve, so it is essential to deposit paychecks (oxygen and glucose) into the bank on a regular basis. To produce the energy (ATP) needed to power cellular functions, all animal cells, including brain cells, use a specialized internal structure called a mitochondrion.
Neurons are greedy consumers of energy for two reasons. First, they have a huge surface area; with projections extending up to a yard, they have a larger surface area than any other cells of the body. Second, neurons must maintain a specialized mixture of ions to power the “battery” that allows them to transmit electrical and chemical signals. These huge energy requirements mean that mitochondria inside the cell are particularly numerous and stressed in neuronal cells. Further, mitochondria are believed to act as calcium sponges, and since neurons are constantly controlling the ﬂow in and out of calcium ions (preferably in small numbers), mitochondria are continually helping to sequester and reuse calcium.
In apoptosis, any one of several triggers sets into motion a cascade of events that seems to affect the mitochondria.4 For example, the cell membrane may allow excess calcium in, thereby shutting down the mitochondria by forcing them to take more calcium than they can handle. Like some oil tankers, mitochondria have double-walled hulls: the outer membrane allows small ions to pass (“permeability”), while the movement of substances through the inner membrane is tightly regulated. In the mitochondrial permeability transition, this delicately balanced system breaks down. The selective permeability of the inner membrane is lost, so the ability to make ATP is severely reduced or destroyed altogether. The cell becomes energy-starved. In the absence of normal metabolism, toxic by-products are generated.
Appropriate apoptosis can be thought of as heroism at the cellular level. A neuron dies so that others may live, or so that the organism as a whole will thrive.
As the mitochondria struggle to maintain their metabolism, the mitochondrial permeability transition also independently sets in motion the cascade of apoptotic cell death. A family of genes, known collectively as caspases, is activated. The genes needed to shut down the cell and elicit a programmed cell death are also activated. Among these are genes called bax, bak, bid, bcl-xS, and the appropriately named bad.
Once these genes go into action, the neuron begins to undergo a characteristic series of changes in form and structure, which reﬂect the underlying molecular changes. The cell nucleus shrinks; DNA is cleaved into fragments; the internal substance of the cell condenses, and some parts are even actively expelled. These changes are all hallmarks of the apoptotic process.
Appropriate apoptosis can be thought of as heroism at the cellular level. A neuron dies so that others may live, or so that the organism as a whole will thrive. Alzheimer’s disease, at least at its most destructive stages, is horribly inappropriate apoptosis.
ALZHEIMER’S DISEASE: CONVERGING ON APOPTOSIS
Alois Alzheimer ﬁrst published his description of characteristic brain lesions in 1911,5 but the concept of Alzheimer’s disease lay dormant, at least for the public, for many years. In the last 25 years, it has ﬁnally been recognized that this neurodegeneration is not a normal result of aging but a disease whose course might be altered.
Alzheimer’s disease can now be diagnosed ﬁrmly only after death, a situation deeply unsatisfying to family members and researchers alike, and useless for the patient. During the autopsy, stained sections of the brain signify Alzheimer’s by the presence of plaques and tangles. Plaques are aggregations of a protein known as amyloid, and tangles are twisted and misshapen bundles produced from the essential structural proteins of dead neurons, the cytoskeleton.
Over the past 100 years, literally dozens of theories have been advanced regarding the cause of Alzheimer’s disease. Originally thought to be a normal part of the aging process, it is now clear that there is nothing normal about the profound memory loss and personality changes that characterize the behavioral aspects of the disorder.
Each theory has its proponents, and this article is not meant to offer support to any of them. It is fair to say that none of them has been able to explain fully all aspects of the disorder, or to give rise to more than minor palliative therapies.
The most popular mechanisms advanced as the cause of Alzheimer’s disease include:
- Loss of neurons that make and release the neurotransmitter acetylcholine (so-called cholinergic neurons).
- Prion diseases. Prions are otherworldly disease agents. Unlike bacteria, fungi, or viruses, they are devoid of DNA or RNA and produce slow, insidious “infections.” Like Kurt Vonnegut’s “ice-nine,” they destroy by acting as templates for the reassembly of proteins surrounding them. The discoverer of prions, Stanley Prusiner, and others have pointed out similarities between Alzheimer’s and known prion diseases such as scrapie, Creutzfeldt-Jakob disease, and kuru.
- Neurotoxicity: aluminum or an excess of other toxic chemicals. Advocates of excitotoxicity claim that the toxic chemicals are produced internally. They hold that the neurotransmitter glutamate, released from neurons or present in excess in the bloodstream, can trigger the death of other neurons.
- ß-amyloid buildup. As noted earlier, one of the characteristics of Alzheimer’s is the presence of plaque, made up primarily of amyloid. It is known that a normal protein is abnormally processed in some neurons, and that the toxic, insoluble protein ß-amyloid is formed. ß-amyloid is excito-toxic, and sets in motion a cascade much like that described for glutamate.
- Tau hyperphosphorylation. The protein tau helps keep certain internal cellular structures in place. Addition of too many phosphates onto tau (hyperphosphorylation) has been shown to lead to the formation of the deranged, wild-looking neuroﬁbrillary tangles that are a hallmark of Alzheimer’s.
- Genetic markers. Different genetic makeup in different individuals has been shown to predispose a person to Alzheimer’s. Among the genetic markers currently under intense investigation are proteins called presenilin and apolipoprotein E.
Obviously, we do not yet fully appreciate the scope of this neurological disorder, we are forced to examine a rich and complicated landscape through a tiny peephole. Whatever the predispositions and the insults and damage along the way, we do know that Alzheimer’s ends in the apoptotic death of neurons in critical centers of the brain.
Understanding what genes are involved in apoptosis, and understanding the functions of the molecular switches that turn apoptosis on or off gives us hope that scientists may learn how to control the process—or simply to disable the switch. We should pause, though, before ﬂipping the switch, to determine whether it might control an essential function.
CAN THE BRAIN REBUILD ITSELF?
Apoptosis was initially understood as natural and necessary. If we could prevent the leaves from falling off the trees every autumn, would we? What might happen if we interfered with the natural order of things? Clearly, autumn is part of a natural cycle that should not be disrupted. This seems obvious to us because we know that in time spring will arrive and the buds, then the leaves, will regenerate.
Unfortunately, for those suffering from diseases that trigger apoptosis, there appears to be no coming spring, no new leaf buds to replace the leaves that have been lost. Twenty years ago, we were taught that humans are born with all the neurons they will ever possess. We now know that there is modest replacement of nerve cells throughout life in many species, including our primate relatives, and it is tempting to think we may be able to replace lost neurons. But with current technology and even foreseeable future technology, we will not be able to regenerate lost neurons in the numbers needed to replace entire systems of the brain. A few million cells might be replaced at a time, but this is a small part indeed of the human brain, which consists of something like 100 billion neurons.
Why doesn’t the brain heal itself? Is there a simple way to explain why no organism has evolved a brain that can self-regenerate?
If we can stimulate neuronal cell division, or replace dying neurons with fetally derived cells, why is it that the brain cannot rebuild itself? I would argue that without a better understanding of the process of development, fetal-brain-cell transplants or stimulated replacement of neurons will be useful only in limited cases. What we dream of is reestablishing lost connections— connections that were formed in an embryo, where the distances to be traveled are small and the intervening tissue is usually a loose, gelatinous meshwork. The axons making the connections were spun out as the target diverged from the nervous system, like a rope extending out behind a harpooned whale. Trying to harpoon that whale through a series of concrete walls when the whale is already miles away would present an entirely different problem.
DOES APOPTOSIS HAVE A FUNCTION OUTSIDE OF DEVELOPMENT?
I toil in these ﬁelds, and each day I hope to ﬁnd out more about why neurons die and how their death can be averted by better control over the molecular and genetic switches that set it in motion. I have some nagging doubts, however. Apoptosis is a natural and, some would say, inevitable process. It has its uses; evolution found apoptosis adaptive for neurons. We need to understand what is “right” about apoptosis before going too far down the path of altering it. Put another way, why doesn’t the brain heal itself? Is there a simple way to explain why no organism has evolved a brain that can self-regenerate?
Consider a similar biological process. In development, certain genes called proto-oncogenes are activated in a speciﬁc sequence to create body structures and organize tissue formation. When activated abnormally later in life, however, protooncogenes cause unregulated cell growth— the array of diseases we call cancer.
Perhaps, like the proto-oncogene mechanism, apoptosis is a latent function. In other words, it is important in development but may be activated only abnormally in adulthood. It is as if there are molecular switches that are necessary in building the nervous system but that should then be covered over and locked, so that they cannot be accessed by mistake at some later point in life. Neurons are exquisitely capable of locking out proto-oncogene proteins: cancers of the neurons themselves are virtually unknown past early childhood. Why, then, are neurons incapable of locking out the apoptotic signal? Is it merely bad luck? I propose an explanation in which apoptosis plays a positive role throughout our lives.
BALANCING PLASTICITY AND SOLIDITY
The nervous system must maintain a balance between plasticity, on the one hand, and something that has no single name— perhaps we could call it stability or solidity —on the other hand. Plasticity is well known to neurobiologists. It is the capacity of the nervous system to adapt itself to changing circumstances. For example, nothing in our hard-wired, genetic programming has speciﬁcally prepared us to drive a car or pilot an airplane. In fact, one could argue that these are abnormal functions. A car moves about 3 meters and an airplane about 30 meters in the one-tenth of a second that the normal human being requires to react to a stimulus.
Plasticity (and some clever human engineering) enables us to overcome these circumstances. Because we are plastic, we can learn to drive. Similarly, we are genetically programmed to see, but even our visual system is subject to plasticity. If we are ﬁtted with goggles that invert our view of the world, we initially become confused and nauseated. In time, though, the visual system adapts to the new circumstances; at that point, removing the goggles would produce the same disoriented feeling as the initial view with them did.
To understand plasticity better, imagine identical twins separated at birth. Both have the same genetic makeup, and therefore are hard-wired for the same functions. If we place one of these twins in a small Iowa town and one in central Beijing, one will learn American English as a native language, while the other will learn Mandarin Chinese. The brain is almost certainly hard-wired for language; but it does not care which one.
Like most human traits, plasticity varies from individual to individual. My capacity to learn languages at the age of 41 is apparently near zero. Contrast this to Joseph Conrad, born in the Ukraine of Polish parents. At 16, he went to sea on French vessels, and by 38 was writing British English ﬂuently as his third language. The language centers of his brain were so plastic that they enabled him to create some of the best-known literature in the English language, even though he did not begin speaking and writing English until he was almost 30.
I want to make the point that the opposite of plasticity, which I will call solidity, is equally critical. Solidity helps us remember what our grandmother’s face looks like, how she smells, where she lives (or lived)—memories that we will retain for a lifetime. At the molecular level, solidity involves a process that in development is called synaptic stabilization, but in adulthood is called long-term potentiation, or LTP.
The process that ensures lifetime learning also holds the potential for dementia and eventual death.
LTP, ﬁrst proposed in 1949 by the Canadian psychologist Donald Hebb,6 7 is a type of connection between neurons. In a “Hebbian synapse,” the more a particular connection is used, the more likely it is to be used again. As a few water droplets follow a particular path running downhill, they deepen a groove, which facilitates the passage of even greater rivulets. In the brain, a few intrepid neurotransmitter molecules clear the way for streams and rivers of consciousness.
LTP is a permanent—solid—change in the chemistry of both partners in the communication: the presynaptic neuron, which sends the information, and the postsynaptic cell, which receives the information. For this synaptic contact to become permanent, a complicated biochemical dance must take place. It begins when the presynaptic cell increases its release of a neurotransmitter (classically, glutamate), which acts on a speciﬁc receptor. This receptor allows calcium to enter the postsynaptic cell, which sets in motion a cascade of events culminating in increased synthesis of nitric oxide. The dance continues, as this gas goes back across the synaptic cleft in the opposite direction from the ﬂow carried by glutamate, making glutamate release more likely and setting up a regenerative cycle. In all likelihood, these synapses continue to be strengthened and solidiﬁed by occasional use that bolsters the biochemical machinery.
Just as there are appropriate and inappropriate modes of cell death, there are appropriate and inappropriate uses for glutamate. Normal neurotransmission and the strengthening of synapses in LTP are appropriate uses. In Alzheimer’s disease, however, ß-amyloid may stimulate a process that triggers glutamate release, and this glutamate release is not helpful. It is acting inappropriately. If a stroke causes glutamate to be freed from dead or dying cells in its core, and glutamate diffuses away to kill the surrounding neurons, their death is inappropriate. In a sense, the neurons have become trapped by plasticity. In this way, the process that ensures lifetime learning also holds the potential for dementia and eventual death
WHAT IF WE COMPLETELY HALTED APOPTOSIS?
That the essential biochemical machinery needed to stabilize our nerve synapses and the machinery needed to trigger apoptosis are virtually the same is an intriguing notion, to say the least. As I have described, synaptic stabilization and LTP depend on glutamate receptors that permit calcium entry and the subsequent production of nitric oxide. But in apoptotic cell death, glutamate-mediated calcium entry and nitrous oxide production are the key triggers that set off the cascade of events we call programmed cell death.
Could apoptosis have a role in the adult brain, mediating an additional form of plasticity? Might the formation of permanent neural circuits depend on apoptosis just as critically as on synaptic stabilization in the early years and LTP in the adult years?
A debate now rages about the number of neurons that can be replaced in the adult brain. For a long time that number was thought to be zero, or so close to zero as to be insigniﬁcant on a background of 100 billion-odd cells in an adult human nervous system. Now it seems that ongoing neuronal division and replacement play a critical role in plasticity in some animals, including primates. Locations with maximum plasticity, such as the area responsible for the learning and production of bird-song, seem to be where it is easiest to demonstrate neuronal cell division and replacement.
Apoptosis might be a part of this process. It could work throughout life to destroy cells that are nonfunctional, inappropriately wired, or no longer needed, so that neurons that are needed can be chosen from the small proliferating population and wired into a network.
For example, a group of neuroscientists is seeking to better understand the plasticity that occurs after nerve cells supplying a limb are damaged or lost. They have shown that the somatosensory cortex—which would normally receive the signals from the limb, process them, and bring them to conscious perception—is reorganized in response to this loss of input (“deafferentation”). Most likely, a failure in the proper reorganization results in the phenomenon of the phantom limb, in which a person feels pain in a limb that is no longer present. Perhaps if we could aid the process of apoptosis in this location, we could increase the plasticity of the appropriate area of somatosensory cortex and thereby relieve the pain.
Current dogma holds that apoptosis is unmitigatedly bad, and many laboratories are working to ﬁnd ways to stop it. Might this be the wrong approach? We must investigate the possibility that by halting apoptosis, we may be removing a mechanism that is critical for the maintenance of plasticity. A more complete view might allow us to control apoptosis regionally or temporally, rather than halting it entirely.
Clearly, a totally solid brain would be a disaster. We could learn nothing beyond the behaviors and abilities inherited as our genetic birthright. Just as clearly, a totally plastic brain would be a disaster. We could retain nothing, and no experience would modify the structure of the brain and store information for future use. As in many aspects of life, a balance between plasticity and solidity is essential. As we come to understand apoptosis, perhaps even learning to manage it in diseases such as Alzheimer’s, we must consider its beneﬁts and uses and take care not to tally losses that will outweigh our gains.