The Mind and the Brain (9 page)

The striatum, and especially the caudate, can thus be thought of as a neuronal mosaic of reason and passion. It sits smack dab at the confluence of messages bearing cognitive content (courtesy of the matrisomes, where inputs arrive from the rational prefrontal cortex) and messages shot through with emotion (thanks to the striosomes, landing zones for inputs from the limbic system). The juxtaposition of striosomes and matrisomes therefore seems highly conducive to interactions between emotion and thought. Since the striosomes receive projections primarily from the emotional centers of the limbic system and the matrisomes receive projections from the higher cognitive centers of the prefrontal cortex, together they provide the perfect mechanism of integrating the messages of the heart with those of the mind.

In the mid-1990s researchers discovered a subset of highly specialized nerve cells that provide a key to understanding how the brain integrates reason and emotion. Called
tonically active neurons
(TANs), these cells often sit where striosomes and matrisomes meet, as Ann Graybiel and her colleagues at the Massachusetts Institute of Technology discovered. The TANs are thus perfectly
positioned to integrate information from both structures and, by implication, from the intensely passionate limbic system and the eminently reasonable prefrontal cortex.

Figure 2:
Cells in the caudate known as tonically active neurons (TANs) tend to be found between striosomes and matrisomes.Striosomes are areas where information from an emotion-processing part of the brain, the amygdala, reaches the caudate; matrisomes are clumps of axon terminals where information from the thinking, reasoning cerebral cortex reaches the caudate. By virtue of their position, TANs can integrate emotion and thought.They fire in a characteristic pattern when the brain senses something with positive or negative emotional meaning. Cognitive-behavioral therapy may change how TANs respond to OCD triggers.

TANs respond dramatically to visual or auditory stimuli that are linked, through behavioral conditioning, to a reward. As a result of this finding, Graybiel’s team began to suspect that TANs play a central role in behavioral responses to hints of an upcoming reward. In a series of experiments on macaque monkeys, the MIT scientists found that TAN firing rates changed when a once-neutral cue became associated with a reward. Let’s say, for instance, that a visual cue such as a flashing light means that the monkey will get a reward (juice) if it performs a simple behavioral response (licking a spoon). When TANs detect a potential reward, that is, they pause at first and then fire faster. But TANs do not respond to the light cue if the monkey has not learned to associate it with a reward. As
the monkey’s brain learns to recognize a reward, TANs fire in a characteristic pattern.

Thanks to its TAN cells, then, the striatum is able to associate rewarded behavior with particular cues. Because TANs can quickly signal a switch in behavioral response depending on the meaning of a stimulus (“That light means juice!”), they may serve as a sort of gating mechanism, redirecting information flow through the striatum during learning. As noted earlier, the entire striatum acts as an automatic transmission: the putamen shifts between motor activities, and the caudate nucleus shifts between thoughts and emotions. Different gating patterns in the striatum may thus play a critical role in establishing patterns of motor as well as cognitive and emotional responses to the environment. Such patterned responses are nothing more than habits. Indeed, Graybiel has shown that the striatum can play a fundamental role in the development of habits. Our best guess is that the tonically active neurons underpin the gating of information through the striatum and thus its role in the formation of habits. What seems to happen is that distinct environmental cues, associated with differing emotional meanings, elicit different behavioral and cognitive responses as TANs shift the output flow of the striatum. In this way TANs may serve as the foundation for the development of new patterns of activity in the striatum.

Most important, TANs could be crucial to the acquisition of new behavioral skills in cognitive-behavioral therapy. In neurological terms, we could say that cognitive-behavioral therapy teaches people purposefully to alter the response contingencies of their own TANs. This is a crucial point. Such therapy teaches people to alter, by force of will, the response habits wired into their brains through TANs. In the case of OCD, therapy teaches patients to reinterpret their environment and apply their will to alter what had been an automatic behavioral response to disturbing feelings. If that happens often enough, then the new response—the new behavioral output—should itself become habitual. The key to a successful
behavioral intervention in OCD, it seemed to me, would be to teach the striatum new gating patterns.

The gating image turns out to be particularly apt in light of what we have learned about the striatum’s two output pathways: one direct and one indirect. The indirect pathway takes the scenic route, running from the striatum through the globus pallidus, to the subthalamic nucleus, back to the globus pallidus, and finally to the thalamus and cortex. The direct pathway runs through the globus pallidus, then straight to the thalamus and back to the cortex. The crucial difference is that the direct pathway provides activating input to the thalamus, but the indirect pathway provides inhibitory input. Thus the direct and indirect output pathways from the striatum have opposite effects. The direct pathway tends to activate the cortex, the indirect pathway tends to quiet the cortex.

The striatal gate determines which pathway nerve impulses will follow. Recall that the striatum receives input from the entire cortex, with the caudate specifically receiving strong input from the prefrontal areas. Prefrontal inputs include those from the orbital frontal cortex and anterior cingulate error-detection circuitry. In 1992 Lew Baxter, my longtime colleague at UCLA, dubbed the circuit containing the orbital frontal cortex and its connections to the basal ganglia the “worry circuit.” It is now often called “the OCD circuit.” When this circuit is working properly, the result is a finely tuned mechanism that can precisely modulate the orbital frontal cortex and anterior cingulate by adjusting the degree to which the thalamus drives these areas. When that modulation is faulty, as it is when OCD acts up, the error detector centered in the orbital frontal cortex and anterior cingulate can be overactivated and thus locked into a pattern of repetitive firing. This triggers an overpowering feeling that something is wrong, accompanied by compulsive attempts somehow to make it right. The malfunction of the OCD circuit that our UCLA group found in OCD patients therefore makes sense. If the exquisite balance of the direct and indirect pathway outputs from the basal ganglia is impaired, it can cause
the orbital frontal cortex to become stuck in the “ERROR! ERROR!” mode.

When the striatum is working normally, it gates the vast array of information about the environment that flows into it from the cortex and initiates what Ann Graybiel has termed “chunks of action repertoires.” These chunks help form “coordinated, sequential motor actions” and develop “streams of thought and motivation.” Thus a single bit of information, such as the feel of a stick shift in your hand, can initiate a complex behavior, for instance, a series of foot movements on the clutch and hand movements on the stick. But in OCD patients the striatum, our PET scans showed, is not even close to functioning normally. It does not gate properly, leading to serious overactivity in the orbital frontal cortex. The intrusive, persistent sense in OCD that something is wrong seems to be the result of orbital frontal cortex neurons’ becoming chronically activated (or inadequately inactivated) as a result of a gating problem, which causes the direct-output pathway to overwhelm the indirect one.

In OCD, the striatum—in particular, the caudate nucleus—appears to be failing to perform its gating function properly. It has become like an automobile transmission that fails to shift. Most people have brains that shift gears automatically, but OCD patients have a sticky manual transmission. As a result, the direct pathway seems stuck in the “on” position. This is what I came to call Brain Lock: the brain can’t move on to the next thought and its related behavior. Instead, such evolutionarily ancient drives as washing and checking for danger keep breaking through, creating a sense of being overwhelmed by these feelings and urges. The feeling of being “stuck in gear,” which often manifests itself as the feeling of needing to get things just right, also explains why an OCD patient finds it so hard to change the compulsive behavior, and why doing so requires such focused and even heroic effort. Medications that block the neuronal reuptake of serotonin can help by at least partially
decreasing the intensity of OCD urges, probably by helping to rebalance the direct and indirect pathways.

A third brain region implicated in OCD is the anterior cingulate gyrus, which also sends projections to the striosomes of the caudate nucleus. Located behind and above the orbital cortex, the cingulate also has connections to the vital brain centers that control the gut and the heart. This structure is probably responsible for generating the gut-churning sense among OCD sufferers that some cataclysm will befall them if they fail to act on their compulsion, say, to tap the steering wheel ten times (or one hundred!) before turning the ignition. The anterior cingulate seems to amplify the gut-level feeling of anxiety and dread.

Even as our UCLA group was working out the OCD circuit, a study from the other side of the country confirmed what we were finding. Researchers at a Massachusetts General Hospital (MGH) group led by Scott Rauch used both PET and functional magnetic resonance imaging (fMRI) scans to measure cerebral blood flow in the brains of eighteen OCD patients. The scientists deliberately created an environment designed to agitate: when a patient settled into the PET scanner, the researchers placed beside him a dirty glove or other OCD-triggering object. The patient’s anxiety level soared. At the same time, the MGH group reported in 1994 and 1996, the PET and fMRI scans consistently picked up significant increases in cerebral activity in the orbital frontal cortex, the anterior cingulate gyrus, and the caudate nucleus—exactly the structures found to be hypermetabolic in our PET scans at UCLA. Looking at the scans, you could almost see the brain desperately emitting “TILT! TILT” messages, signaling that something was dreadfully wrong. The conclusion was clear: when OCD urges become more intense as a result of exposure to a trigger such as a dirty object, circuitry involving these three brain structures becomes more active.

A picture of the brain abnormalities underlying OCD was emerging. The malfunctions center on circuitry within the orbital frontal cortex, containing “error alarm” circuits, and the basal ganglia, which acts as an automatic transmission or switching station. The circuit responsible for detecting when something is amiss in the environment, centered on the orbital frontal cortex, becomes inappropriately and chronically activated in OCD, probably because a malfunction in the gating function of the caudate nucleus allows the prefrontal cortex to be stimulated continuously. The result is a pattern of intrusive, persistent thoughts and feelings that something is wrong or excessively risky. Interconnections among the orbital prefrontal cortex, anterior cingulate, and caudate may allow this circuit to become self-sustaining and thus extremely difficult to break out of—as any OCD patient can attest. The result is a perseveration of the thoughts and urges that OCD creates. That these abnormalities leave the superior prefrontal regions, and thus higher cognitive function, essentially untouched seems consistent with the ego-dystonic nature of OCD’s intrusive thoughts and urges—that is, with the fact that they are experienced as alien to the patient’s sense of who she is, and apart from the natural flow of her stream of consciousness.

This neurobiologically based view of OCD did not exactly take psychiatry by storm. At a meeting on anxiety disorders in the early 1990s, I was presenting some recent findings, on a poster, about the brain mechanisms underlying OCD. A leading behavioral therapist strolled up to me, stopped, and glanced at the poster. Looking me up and down, she spit out, “The whole notion you have of the brain causing OCD is ridiculous!” Well, I asked, what else might be the cause, if not the brain? “I’ll tell you what causes OCD. Try not to think of a pink elephant. Can you do it?” Without waiting for an answer, she continued, “There! That is what causes OCD”—and she walked away.

Fortunately, responses like this were the exception rather than the norm, and the overheated brain circuitry underlying OCD, as we and others were working out, offered a glimmer of hope for patients. Since the most evolutionarily recent (and thus sophisticated) prefrontal parts of the brain are almost entirely spared in OCD, the patient’s core reasoning power and sense of identity remain largely intact. They might thus be relied on, I reasoned, to play a part in therapy. Now that science had pretty much nailed down the brain abnormalities at the root of OCD, I was ready for the next step. I set out to find a treatment that would alter metabolic activity in the guilty triad: the orbital prefrontal cortex, the anterior cingulate gyrus, and the caudate nucleus. In particular, I had a hunch that any successful treatment would probably have to enhance the gating function of the caudate so that the worry circuit could be quieted and the patient enabled to resist OCD urges.