The Brain-Immunology Axis

Of all biological systems, the most complex, by far, are the human brain and human immune system. Each, in its distinctive way, has evolved the capacity to remember experiences, learn, and communicate. But how do they talk to each other? Do the two systems share a common tongue? If they argue, who wins?

Using approaches such as studying autoimmune diseases of the nervous system, when the immune system mistakenly attacks the body’s own cells, as in multiple sclerosis (MS), scientists are bringing into focus the delicate dance of communication between neurons and immune cells. Understanding how immune processes work in the nervous system itself could lead to breakthrough treatments for conditions ranging from spinal injury and stroke to Alzheimer’s disease and Parkinson’s disease. Although we have learned that psychological experiences like stress affect our vulnerability to illness, neuroimmunology can pin down how that process actually takes place and how we can increase our resistance to it.

This may be only the beginning. As we discover what regulates immune power and whether our brains have voluntary control over it, medical science could take possession of a tool of enormously diverse potential. Neuroimmunology has implications for lifelong health, disease prevention, and even defense against bioterrorism.

Round One: Discovering the Connections

Although the immune system has been investigated intensively since Louis Pasteur created one of modern medicine’s marvels, vaccination, 150 years ago, the study of brain-immune system interaction is new. Immunologists did not see the need for it. True, the brain is the body’s command center, but the immune system—unlike our organs—is not in one place. Immune cells rove freely through the blood and into most of our tissues and can respond relatively independently to local conditions. Bruce S. McEwen, Ph.D., author of the forthcoming The End of Stress As We Know It (Dana Press/Joseph Henry Press, 2002), says: “The immune system has its own rules for operation, which it can do perfectly well in a dish.”

Because guidance from the brain seemed unnecessary for immune response, and because immunity is so complex, attempts at simplification ruled most of 20th-century immunology. Immunologists felt free to claim that neural input was not important, even if it did exist, thus simplifying their field of study. They already had their hands full working out the roles of thousands of chemical messengers and at least 62 different types of cells involved in immune reactions, each sensitive to minute chemical changes and often responding differently to the same stimuli at different times or in different locations.

Yet, as early as the 1800s, anatomists had recognized that immune system organs like the thymus, bone marrow, and spleen, where immune cells grow and differentiate, and lymph nodes, where immune cells drag their prey, had more connections to the brain than their size or location alone would seem to require. These suggestive findings were apparently first ignored and then forgotten. But working in McEwen’s lab in the 1970s, Karen Bulloch, then a Ph.D.-degree candidate, rediscovered these data and traced the nerve connections to major immune organs, using modern techniques to confirm and update the old observations, which are now generally accepted.

Over the past decade the need for a specialized field focusing on the links between neuroscience and immunology became evident, then urgent. Scientists began to find receptors on T-cells—key immune soldiers that learn to recognize and attack a specific invader every time it enters the body—and discovered that the receptors bind to neurotransmitters such as endogenous opiates and norepinephrine. Clearly, the immune cells were equipped to receive messages from the brain. Scientists also spotted traffic flowing in the other direction: immune chemical messengers, called cytokines, could affect the brain.

Then the links became even closer. Consider the simultaneous search by neuroscientists for the cause of fever—the “endogenous pyrogen”—and the quest by immunologists for the “leukocyte activating factor,” a chemical that turns on the immune system’s white blood cells that multiply when we are sick. As it turned out, they were searching for the same molecule: interleukin-1 (IL1). “That was a real eye opener, that the two [systems] use common messengers,” says Tamas Bartfai, director of the Neurological Research Center at Scripps Research Institute. Additionally, researchers recently found that substances called neurotrophins, including nerve growth factor, which are known to mediate survival of nerve cells, also appear to affect survival of immune system cells.

Earlier in the 20th century, researchers had recognized that glucocorticoids—a family of steroid hormones produced by our brain and endocrine system when we are under stress—could seriously affect immune responses. In 1950, the Nobel Prize in Medicine or Physiology went to three scientists who confirmed that these hormones suppress inflammation. When used as drugs, glucocorticoids like prednisone and cortisone can slow down the immune system or even entirely turn off specific elements of it, a useful outcome in areas of medicine from allergy and asthma treatment to prevention of transplant rejection.

Immunologists, though recognizing that stress hormones had the ability to switch off the immune system, first concluded that this would take place only when hormones were present at unnaturally high, pharmacological doses, not under normal circumstances. After all, it made no evolutionary sense for an animal under stress to use hormones to turn off its immune response at the moment it might need it most.

When immunologists found that stress hormones could affect immunity under normal bodily conditions, not just when influenced by medications, one hypothesis was that the brain’s hypothalamic-pituitary-axis (HPA), which controls the output of stress hormones, also controls the immune system. If this were the case, perhaps the brain and immune system were not affecting each other; perhaps the brain was generating both the stress and immune responses.

“The discovery of cytokines changed this,” says Bartfai. For example, IL1, produced by immune cells in the body when it is fighting infection, can have profound effects on the brain by helping cause what researchers call “sickness behavior.”

“I’m an example right now of how immune cytokines affect the brain,” said Esther Sternberg, Ph.D., director of the integrative neural immune program at the National Institute of Mental Health, sounding hoarse from a cold when interviewed for this article. “Sickness behavior involves functional changes in mood, memory, cognition, and sleep, and it also activates the HPA stress response.” In other words, when you are sick, it is your immune system that is telling your brain that you should lie down and take it easy. Whether you respond or not may have as much to do with the strength of the signal as with the strength of your will to keep going; no one knows.

“It’s certainly a two-way interaction, where the immune system signals the brain and the brain signals the immune system and the two, in concert, act as a rapid response system of the body to all sorts of external stimuli,” says Sternberg.

“The immune system can be viewed as a sensory organ,” she adds, “sending signals about pathogens [just as] the eyes send visual signals and the ears send auditory signals. The brain responds and produces hormones and neurochemicals that alter immune function.”

Seeking a more detailed picture of the brain’s stress response, neurobiologists asked how and why stress hormones released by the adrenals in response to signals from the brain could affect immune reactions. Here they encountered another obstacle: Most scientists were skeptical of any suggestion that the “mind” could affect health—an area of research recently dubbed “neuropsychoimmunology”—because that sounded too much like the claims of those who thought positive affirmations could cure cancer. Summing up this era, Richard Ulevitch, Ph.D., chairman of immunology at the Scripps Research Institute, says: “Neuroimmunology is rising, now. But it hasn’t been very well respected in the past.” He attributes this to the subject’s complexity as much as to skepticism about the ability of the mind to control immune function: “The three most complex biological issues are neurobiology, immunology, and endocrinology. Each is a tough subject on its own. Most people who finally master one of them have no desire to master another.”

Complexity confounded much early research. For example, studies of how psychological stress affects health reach notoriously conflicting conclusions; some find that stress matters, some that it does not. Even when it is shown to matter, the size of the supposed effect differs greatly. In yet another puzzle, if stress turns off the immune system, then why do people with autoimmune diseases like rheumatoid arthritis or MS seem to get worse, not better, during stressful periods? If their symptoms are caused by their immune system’s attack on their own bodies, shutting down immunity as a result of stress should help.

Neuroimmunology is now clarifying the picture. Almost everyone has experienced a stressful period before an important exam or deadline, gotten through the big event, and immediately fallen ill with a flu or cold. What stress does to the immune system, neuroimmunologists now believe, is enhance it during short-term crises and diminish it if the stress persists and becomes chronic. McEwen says: “It helps under acute stress to keep you going and then you fall apart afterward.” This seems to square better with evolution and explain why stress research has been so difficult to interpret. 

Round Two: Who’s in Charge Here?

Figuring out what controls the interrelationship between these two systems is hard. What we see is not a top-down command system but a series of interconnected loops, influenced by reactions to changing conditions. Trying to identify the leader is like trying to deduce who is leading a jazz band without knowing the relationships between any of the players or the conventions of the music. There is certainly order and the musicians clearly signal each other somehow—but how, and who is making the ultimate decisions about where the improvisation will go, at times may not be clear even to the players themselves.

Richard M. Ransohoff, M.D., a neuroscientist at the Cleveland Clinic Foundation, says: “Strong scientific evidence that the brain controls the immune system just isn’t there.” He asks: “Do we really have to add another level of stress, so that people suffering from diseases have to worry that they aren’t relaxed or positive enough?”

Certainly the brain can modulate some aspects of immune operation, but this does not necessarily imply that we are in conscious control of our immunity, any more than the brain being able to mediate heart rate implies that we can do so at will. But can control over immune response in any sense be learned? That is what scientists call “an elegant question.”

Research has suggested ways in which psychological factors that are under our control (for example, having a support network of family and friends) can improve our health by relieving stress, but it is rife with contradictions. For example, a study published in the New England Journal of Medicine in December 2001 by P. J. Goodwin and his colleagues at the University of Toronto failed to confirm earlier work suggesting that women with breast cancer who attend support groups live longer.

These intriguing questions about the specific means by which voluntary processes may influence immune response are a scientific frontier of potentially enormous importance, but are also highly complex. Focusing on brain and immune processes that are involuntary, and also complicated, has yielded solid evidence of interaction. 

When Things Go Wrong

One way to look at involuntary processes involved in brain/immune interaction, and what controls these interactions, is to examine autoimmune diseases. What happens when the interaction goes wrong, as it does in autoimmune diseases such as MS, rheumatoid arthritis, juvenile diabetes, and lupus? These are very different diseases that produce vastly different problems, but they seem to have a common problem, as well: Immune cells, for reasons we do not yet understand, misidentify certain of the body’s own cells as foreign and attack them. There are risks in extrapolating from sickness to health: The illness itself may change crucial factors. But scientists studying MS, for instance, are shedding new light on key neuroimmune connections that may be at work in many disorders.

Investigators now suspect that MS is not a single disease and that the symptoms of MS may arise from any one of a small group of similar patterns of tissue pathology in the brain and spinal cord. Depending on which nerves are affected, people with MS have symptoms ranging from tingling sensations and dizziness to incontinence, vision loss, and paralysis. The common underlying process seems to be that immune cells attack the insulating covering (myelin) of neuron extensions, called axons, profoundly affecting the nerve cells’ ability to communicate. The axon is like a cable through which a neuron sends its messages over long distances; remove the cable’s protective coating and the long-distance service will provide increasingly bad connections.

In some cases of MS, rogue immune cells launch the attack; in others, defects in the neural wiring or the protective myelin cover may attract the negative immune attention. Both problems may have to be present to some extent for the disease to take hold, which some scientists think happens only when a susceptible person is exposed to a virus that alters myelin to look foreign to the immune cells. But how might two seemingly unrelated problems—rogue cells and tissue that attracts them—occur simultaneously in some unfortunate people? Four out of five of us have some cells capable of attacking myelin, but fewer than one in a hundred of us will develop MS. The reason is that autoreactive cells, which identify the body’s own cells as foreign, are generated quite readily but are very difficult to activate, a process that we do not yet understand.

One hypothesis is that some of the same factors shape development of both the nervous and immune systems. An example is the genetic factor in MS. People who inherit a certain variety of genetically determined immune-system markers, called HLA haplotypes, are at significantly greater risk for MS. “It is not clear how these confer susceptibility,” says Roland Martin, Ph.D., acting chief of cellular immunology at the National Institute of Neurological Disorders and Stroke, but one clue may be that a person’s HLA haplotype affects not only the immune system (the repertoire of T-cells in the thymus and how well the body rids itself of defective immune cells prone to attacking the body’s own cells) but also the nervous system (the wiring of certain brain areas). The person’s HLA haplotype may be making those brain areas more susceptible to immune attack.

When Immune Cells Inflict “Collateral Damage”

While many scientists are studying the process of myelin damage by immune cells, Michael Carrithers, Charles Janeway, and their group at Yale University want to know how destructive immune cells are attracted, and gain access, to the central nervous system, where they do their dirty work. If these cells slip past the usually hard-topenetrate blood-brain barrier, that barrier starts to break down and a chain of devastating events, which we do not know how to reverse, is set in motion. Finding out how this occurs may be crucial not only to treating MS but to our understanding of neuroimmunology itself.

Until recently, scientists thought that the blood-brain barrier, which is made up of special cells and enzymes that bar most molecules in the body’s primary blood flow from the interior of the brain and central nervous system, rarely if ever admitted immune cells from the periphery of the body into the tightly regulated space around the brain and spinal cord. Now, they are discovering in the brain both specialized immune cells native to the nervous system and T-cell lymphocytes from the periphery that do gain entry and patrol for infectious agents. “Even in the absence of disease, there are always T-cells circulating throughout the brain,” says Carrithers. His group is charting how the brain and immune system regulate the number and types of T-cells admitted and how immune surveillance of the brain is maintained.

Normally, T-cells are concentrated in the cerebrospinal fluid (CSF) in the meninges, the spaces in the outer lining of the brain. In autoimmune diseases like MS, however, these T-cells leave the fluid and attack brain tissue. How? Carrithers explains that, preceding the onset of MS, changes can be seen in adhesion molecules. These attract immune cells to the blood vessel walls and tell them when to leave circulation and head for the area they are designed to attack (under normal circumstances, an area where an infection requires control). In MS, things go wrong; in an apparent case of mistaken identity, the T-cells attack innocent brain tissue.

“For some reason, brain tissue attracts many more cells in susceptible people,” says Carrithers. “The brain is communicating with the immune system to attract cells, but once the cells get there, they change the local environment and more cells come in.” These cells then attract more cells, and so on. The damaging cascade of events begins.

“We hypothesize that, even in the absence of inflammation, if cells are going through the brain, this alters it and maintains some kind of [balance],” says Carrithers. This balance could keep a steady number of immune cells available in the brain, which would change to meet specific challenges. Increased expression of the adhesion molecule called P-selectin seems to be important in initiating the process of attracting the immune cells that cause the neural lesions seen in MS. Preventing this from occurring, then, could potentially arrest the disease. 

When the Nervous System Regulates Inflammation

MS provides a clear example of how the nervous system can regulate immune cell inflammation. Immune cells are distributed throughout the body, and travel to areas where they are needed to fight infection, but where and how do they travel? Ransohoff works with chemokines, an important subset of immune messenger chemicals called cytokines. He says that “chemokines act on a distinct family of receptors. The practical difference between chemokine receptors and other cytokine or growth factor receptors is that chemokine receptors are among the very few druggable receptors in the immune system.”

By this he means that, unlike other immune receptors, which can be activated or blocked only by using high-tech laboratory molecules that could never become cheap generic drugs, the receptors for chemokines respond to the action of drugs already being developed. Another example of a “druggable” receptor found on both immune and nerve cells is the beta-receptor for the neurotransmitter norepinephrine. Drugs such as propranalol, which block this receptor, are widely used to treat high blood pressure, cardiac arrhythmia, and migraine. It remains uncertain whether such drugs could be used to modulate immune responses.

Ransohoff is studying the mechanisms of inflammation. His work rests on the observation that chemokines and their receptors are involved in highly specific kinds of immune cell traffic, but they have a wide range of actions. According to Ransohoff: “They are involved in a [white blood cell’s] decision to stop flowing in the blood and go into tissue. They can activate an adhesion molecule on the leukocyte and cause it to stop at a particular place and re-arrange its cytoskeleton so that it can crawl between [tissue] cells.”

Chemokines also influence how the immune system matures. “This is relevant to MS because the inflammatory lesions in some way recapitulate the development of the immune system,” says Ransohoff. In other words, the same patterns of recruitment of cells are seen in both development and in the initiation and growth of an MS lesion. “In MS, as best we understand it, T-cells orchestrate an attack on myelin. The T-cells recruit a bunch of macrophages, [immune] cells that consume pathogens and damaged tissue, and direct them to eat the myelin. Once the process is initiated, it is self-perpetuating because you have essentially the formation of a new lymphoid organ in the brain consisting of T-cells, macrophages, and a rich soup of antigens [the proteins that the immune system is attacking].” This is a clear example of how the nervous system talks to the immune system by regulating the process of inflammation using chemokines. Unlike other organs, the central nervous system cannot tolerate swelling. “A direct contrast would be nasal passages. I can accumulate many ‘boogers,’” he says, smiling, “and it’s fine. There’s a virulent inflammatory reaction with huge cell death and production of pus, and it all goes back to normal as soon as I clear the virus.”

But in the central nervous system, the same amount of swelling and fluids could be severely destructive or even deadly. So one way of looking at neuroimmunology is as a field that examines how the immune system is specialized to communicate with the brain and spinal cord to execute ordinary immune operations, while working within the limitations of these delicate tissues. Understanding this could lead to new treatments for spinal cord injuries, brain injuries, stroke, Parkinson’s disease, Huntington’s disease, Alzheimer’s disease, and other types of illnesses that involve brain or neural inflammation gone awry. In acute injuries like spinal cord damage, stroke, and head injury, most of a patient’s lasting problems result not from the initial injury but from the effects of increased inflammation around the injury. In degenerative diseases like Parkinson’s and Alzheimer’s, inflammation kills brain cells. 

New Strategies for Spinal Injury

Linda Noble, professor and vice chairman of neurological surgery at the University of California, San Francisco, explains that controlling inflammation could have a major effect on recovery from spinal cord injury. “What makes spinal cord injuries interesting to study, and where there’s a clinical opportunity for patients, is that a substantial amount of the functional loss that occurs can be prevented. We believe that’s because what causes a generous portion of the damage is cellular reactions afterward, not the injury itself.”

Research in cats shows that only 20 percent of spinal cord neurons have to function for a cat to walk again after injury. That suggests that many spinal-cord injuries that now lead to paralysis might not, if even some of the inflammation could be stopped quickly. By contrast, anti-inflammatory treatment now to help someone like Christopher Reeve, who was paralyzed from the neck down years ago in a riding accident, would not be effective since the inflammation did its damage back then.

Emergency treatment of a spinal cord injury usually calls for administering large doses of steroids to reduce inflammation, but physicians are asking whether this actually contributes to better functioning later, as studies originally had suggested. Noble is exploring other strategies to attenuate spinal cord damage. Here we return to the question of how and where immune cells travel through the body. Immune cells called neutrophils depend on chemicals called matrix metalloproteins (MMPs) when they have to leave the bloodstream and pass through tissue. This is because the MMPs can degrade the outer tissue that surrounds cells—a necessary step, in normal circumstances, that enables a cell to move through solid tissue to get where it needs to go. But MMPs can be dangerous if turned on at the wrong time, as happens in invasive tumors. A neutrophil rushing to the site of a spinal injury generates MMPs and, says Noble, “like a miner, it tunnels its way through. By getting rid of that function, [a neutrophil] can’t very effectively get in.”

A substance called GM6001 can block MMPs. This has been shown to improve outcomes after spinal injuries in animals, if administered within hours after the injury. “There was attenuated neutrophil infiltration,” says Noble. “But of greater importance, if you looked at the animal afterward, it showed significant locomotor recovery.”

The Depression Connection

Inflammation also may be associated with depression when the immune system is activated. Rolipram, initially used as an antidepressant, blocks an enzyme that is available in both the brain and the immune system. According to Roland Martin, blocking this brain enzyme also seems to escalate production of nerve growth factors. Rolipram was used as an antidepressant in Europe and Japan, but its manufacturer decided against seeking approval for this use in the United States because the drug did not demonstrate advantages over existing drugs and its patent has lapsed. Martin has reformulated the drug, which is now in Phase II trials to test it for efficacy in people with MS.

How does depression fit into the picture? Some scientists are finding a link between depression and inflammation caused by immune activation. One immune cytokine, alpha-interferon, which occurs naturally in the body, is known to cause depression as a side effect when given as a pharmaceutical treatment for hepatitis C or other conditions. Depression is also linked with chronically elevated levels of stress hormones. McEwen’s lab has demonstrated how chronically high levels of stress hormones can damage part of the hippocampus, a process that may be responsible for some memory problems associated with depressive illness.

Even more intriguing, all antidepressants seem to work in part by inducing an increase in nerve growth factors, particularly in the hippocampus. That increase might mitigate or repair the damage caused by depression and thereby clarify a puzzle about the action of antidepressants: Although they rapidly raise levels of neurotransmitters believed to be out of balance in depression, the patient’s psychological symptoms will not improve for several weeks. As it happens, several weeks are needed for the increase in nerve growth factors to occur. Could this increase be what actually leads to recovery?

The brain and immune system also seem to be involved in the critical interaction among inflammation, heart disease, and depression. We now know that depressive symptoms are associated with an almost eightfold increase in mortality after a heart attack. But scientists at the University of Pennsylvania found that depressed patients treated after a heart attack with selective serotonin reuptake inhibitors (SSRI) such as Prozac or Effexor had a 35 percent lower risk of a subsequent heart attack. Perhaps antidepressants work by directly or indirectly causing nerve growth factors to reduce the inflammation of heart tissue—a critical element in coronary disease. These drugs could also work through receptors on certain immune cells for the neurotransmitter serotonin. Since serotonin can thin blood by action on platelets, it might reduce the risk of heart attack directly, as well.

Two Very Different Kinds of Memory

One capacity in which the immune system and the brain might be expected to be similar, but turn out to be extremely different, is memory. Both systems must remember relevant stimuli and learn to respond if those stimuli are encountered again. Their processes for memory storage and recall, however, are vastly different. So far, we understand immune memory better: Specific cells (T and B cells) are designed to respond to specific pathogens and change in certain ways once they have encountered their foes. Then, if the foe is seen again, these cells are ready to respond step by step to prevent the same pathogen from causing a second round of illness.

Brain memory is far more complex and flexible. No one knows yet exactly how a memory is stored, but it seems to involve changes in the strength of connections between neurons. In contrast with the immune system, the brain’s recall is not automatic; even if cued by the same stimuli, memory can alter with circumstances. Also, a panoply of sensations and experiences, a whole life’s worth—not just encounters with specifically recognizable pathogens—can be encoded in our memory.

Says McEwen: “The immune system and brain are similar in the chemical messengers they use. But people who try to extrapolate between immune cell memory and human [brain-based] memory are barking up the wrong tree. They are analogous, not homologous.” Furthermore, a neurotransmitter may say one thing to the brain, but carry a very different message in the immune system. Evolution’s tendency to use old structures and chemicals for new purposes to the extent possible has led to a high degree of efficiency in biological systems, but to big headaches for scientists trying to puzzle out what chemical plays what role, where, and whether this role changes depending on local circumstances.

As the new neuroimmunologists probe the web of connections between the brain and the immune system—understanding their commonalities and decoding their messages—the apparent independence of these connections is likely to keep diminishing. There are few more daunting, or more important, challenges in medicine.