Point to any one organ in the body, and doctors can tell you something about what it does and what happens if that organ is injured by accident or disease or is removed by surgery—whether it be the pituitary gland, the kidney or the inner ear. Yet like the blank spots on maps of Central Africa from the mid-19th century, there are structures whose functions remain unknown despite whole-brain imaging, electroencephalographic recordings that monitor the brain's cacophony of electrical signals and other advanced tools of the 21st century.

Consider the claustrum. It is a thin, irregular sheet of cells, tucked below the neocortex, the gray matter that allows us to see, hear, reason, think and remember. It is surrounded on all sides by white matter—the tracts, or wire bundles, that interconnect cortical regions with one another and with other brain regions. The claustra—for there are two of them, one on the left side of the brain and one on the right—lie below the general region of the insular cortex, underneath the temples, just above the ears. They assume a long, thin wisp of a shape that is easily overlooked when inspecting the topography of a brain image.

Advanced brain-imaging techniques that look at the white matter fibers coursing to and from the claustrum reveal that it is a neural Grand Central Station. Almost every region of the cortex sends fibers to the claustrum. These connections are reciprocated by other fibers that extend back from the claustrum to the originating cortical region. Neuroanatomical studies in mice and rats reveal a unique asymmetry—each claustrum receives input from both cortical hemispheres but only projects back to the overlying cortex on the same side. Whether or not this is true in people is not known. Curiouser and curiouser, as Alice would have said.

Unlike most other parts of the brain, there are no reliable case studies of patients with selective destruction of one or both claustra from stroke, viral infection or other calamity. Lesioning the structure in laboratory animals is challenging given its thin and elongated nature. For the same reason, brain imaging has not been very useful: the smallest spatial features distinguishable through positron-emission tomography or functional MRI, two of the most widely used imaging techniques, are two to three millimeters across, bigger than the claustrum's width. And because it is embedded within white matter and sandwiched between two very active neuronal tissues—below the neocortex and above the putamen, part of a larger region, the basal ganglia, lodged deep within the brain—it is problematic to unambiguously pinpoint changes in blood flow to the claustrum and not to these nearby, large structures.

Enter the Dragon

anding function is to study structure. Francis Crick and James Watson proved this idea spectacularly in 1953. They inferred the key function of DNA, the molecule of heredity—that is to say, storing and copying genetic information— from its double-helical chemical structure. Half a century later Crick, by then biology's most respected sage, tried his hand at the same game, linking a structure—the claustrum—to a function—the emergence of integrated, conscious experience.

Whereas scholars of consciousness disagree about many aspects of this most mysterious phenomenon, virtually all agree that one of the defining properties of any subjective experience is that it is unified. No experience can be reduced to independent components. Every experience is irreducible. When I look at my wife's face, I do not see two eyes in a black-and-white picture with a disembodied layer of blue superimposed on top. No, I perceive her blue eyes as one integral and seamless whole. Nor do I experience my Bernese mountain dog doing funny things with her snout while a loud noise fills the room; no, I hear her bark. The experience of seeing the word “honeymoon” is not reducible to the experience of seeing “honey” on the left and “moon” on the right.

We know that different groups of neurons become active in response to such commonly encountered features as colors and motion, faces and dogs, words, sounds, and so on. These cells are dispersed among the 16 billion neurons making up the cerebral cortex. Together the active and inactive cells give rise to a conscious experience. Furthermore, we know from introspection that what we are conscious of is in constant flux.

Distracted by the sight of a passing motorboat on the lake outside my house, I am about to turn back to writing my article when I suddenly recall that I promised to pick up dog food, and then my attention shifts without warning to Richard Wagner's “Liebestod” playing on the radio. Each of these sights, sounds, memories or thoughts requires that the underlying electrical and chemical activity of a privileged set of neurons is rapidly bound to give rise to an integrated conscious experience that lasts but a fleeting moment until the next neuronal assembly comes into being and a new experience supersedes the old one.

Looking at the far-flung two-way connections between the claustrum and the cortex, Crick and I—for at that time in 2004, I was working closely with him and had been for 16 years—hypothesized that this superhub of neuronal activity could be pivotal for consciousness. Because every region of cortex projected to its associated claustral target area, and this neural communications hub reciprocated the connection, the claustrum could serve as an integrator for crisscrossing electrical signals, provided that all of this information could be freely admixed within the structure. We endlessly discussed various neuroanatomical and biophysical means for the claustrum to achieve this integration and wrote a manuscript.

Francis knew that he only had a limited amount of time left; he had end-stage colon cancer. He called me on the way to the hospital, calmly telling me not to worry about the manuscript following our last brainstorming session because he was going to make corrections to it (which he did, dictating them to his secretary from the clinic). Two days later, on his deathbed, Francis hallucinated a debate with me about the role of the claustrum's connection to consciousness, a scientist to the very end. The paper was published a year later in the world's oldest scientific journal, the Philosophical Transactions of the Royal Society.

Enter the Electrodes

In the intervening years, a handful of studies further delineated the molecular neuroanatomy of the claustrum in rodents and a crude map of its connections in people. One investigation focused on the role of the claustrum in integrating visual and auditory stimuli. Using microelectrodes that recorded the electrical activity in awake monkeys, the investigators confirmed that part of the claustrum tended to respond more to visual stimuli, whereas one of its nearby regions was sensitive to tones. But no individual neurons responded to both visual and auditory events, arguing against a multisensory role for the claustrum, thereby leaving it bereft of any obvious function.

This seeming impasse may have changed with a single dramatic case report. A 54-year-old woman who had uncontrollable epileptic seizures had electrodes implanted deep within her brain to help pinpoint the exact origin of her seizures. During this procedure, electrodes can triangulate the focal area where the seizure originates so that it can be surgically removed. They can also inject electric current to help map the brain, identifying areas responsible for important functions such as speech or movement and thus sparing them during the surgery.

Led by Mohamad Z. Koubeissi, an associate professor in the department of neurology at George Washington University, the clinical team made a remarkable observation: electrically stimulating a single site with a fairly large current abruptly impaired consciousness in 10 out of 10 trials—the patient stared blankly ahead, became unresponsive to commands and stopped reading. As soon as the stimulation stopped, consciousness returned, without the patient recalling any events during the period when she was out. Note that she did not become unconscious in the usual sense, because she could still continue to carry out simple behaviors for a few seconds if these were initiated before the stimulation started—behaviors such as making repetitive tongue or hand movements or repeating a word. Koubeissi was careful to monitor electrical activity throughout her brain to confirm that episodes of loss of consciousness did not accompany a seizure.

Two aspects of this patient's case had never been seen before. First, no abrupt and specific cessation and resumption of consciousness have previously been reported, despite decades of electrically stimulating the forebrain of awake patients in the operating room. Depending on the location of the stimulating electrode, patients usually do not feel anything in particular. Less frequently, a patient may report flashes of light, smells or some difficult-to-verbalize body feelings, or perhaps even a specific memory from long ago that the electric current evokes. Or the patient will twitch a finger or a muscle. But this case was different. Here consciousness as a whole appeared to be turned off and then on again. Second, it happened only at a single place, in the white matter close to the claustrum and the cortex. Because electrical stimulation of the nearby insula is not known to elicit a loss of consciousness, the researchers implicated the claustrum.

It is difficult to be confident of the actual causal mechanisms—the stimulation may have triggered electrical discharges from neurons' wirelike extensions to exert effects at another site. Unfortunately, this tantalizing case report cannot easily be followed up with more experiments, because the patient's electrodes were subsequently removed.

We do not have the luxury of waiting for an analogous finding, perhaps as long as a century hence, so it is important to devise experiments to confirm the existence and properties of any claustrum on/off switch. The most promising idea would take advantage of proteins specifically expressed in cells in the claustrum but not in other brain structures. Knowledge of these cells' molecular zip code can then be exploited by tools of molecular biology to quickly and transiently turn the electrical activity of neurons in the claustrum off and on with beams of colored light and to observe the effects on the behavior of lab mice.

If the claustrum truly plays a critical role in generating conscious experiences, we will find out and take another small step toward the ultimate goal of identifying the footprints of consciousness in highly excitable matter. Per claustra ad astra!