Conference 2002

Workshop I
Workshop II
Workshop III
Workshop IV
Final Group Discussion and Summary
Schedule

Photos of the June 2002 Meeting

Final Group Discussion and Summary
Chairs: D. Weinberger and A. Hariri; Scribe: S. Sesack

Three Questions

A. Does abnormal responsivity to stress lead to an increase in susceptibility to neuropsychiatric illness, including addiction, and alterations in paralimbic and limbic regions?

B. What experiments would test the idea that a common neural substrate mediates the appetitive and aversive properties of a stimulus?

C. What kind of experiments could examine affective switching and its basis?

Most of the discussion focused on the first question, with some on the third. Little time was spent talking about the second question.

Regarding the first question, Susan Sesack asked whether stressed animals will self-administer more cocaine than non-stressed animals, to which Michela Marinelli replied that there are studies showing that they do. Isolation stress also increases drug intake. However, this increased drug taking does not, in itself, constitute addiction.

The subject of stress-induced relapse of schizophrenia was raised, to which Daniel Weinberger observed that stress almost never compensates for any condition. Rather, it almost always degrades or depreciates function. Moreover, stress not only worsens conditions but may also cause the emergence of new pathology.

Tony Grace described work of Janet Finlay in which PFC lesions resulted in an exaggerated stress response as measured by dopamine release in the nucleus accumbens. The increase was not only greater than in controls, it also lasted longer. Tony then suggested that a useful definition of a maladaptive stress response would be one that is no longer time limited but persists beyond the originating stimulus.

Bita Moghaddam observed that genetic predisposition may lead to maladaptive stress responses in some individuals. For example, hypoactivity in the orbitofrontal cortex may precede addiction and cause maladaptive responses to stress - not just a stress-induced exacerbation of the condition but an abnormal response altogether. This idea is reminiscent of the role of temperament that was discussed by Jack Gorman earlier in the meeting.

Tony Grace wondered whether anything might be learned regarding this concept by considering patients with post-traumatic stress disorder. Not all individuals with the same stress exposure develop PTSD. Such patients may be predisposed to respond to stress in this way either by temperament or by prior stress exposure before the trauma (i.e., more "hits").

Weinberger noted that susceptibility factors are not always obvious. For example, the actions of alcohol dehydrogenase are associated with alcoholism because of the effect of the enzyme at the level of metabolism the first time someone is exposed to alcohol. In this case, a particular enzyme form predisposes a person to alcoholism, rather than the brain’s response or an "addictive personality."

The question was then asked about the capacity of other systems to adapt to stress. There may be another component of the system that fails to respond normally to stress. Someone noted that there is substantial evidence from various mouse lines for genetic predispositions toward certain behaviors or stress responses. It was then stated that it would be useful to have a model that specifically links genetic background to experience in early development. What alterations in different brain regions might be the most critical? Hypofrontality? Lack of amygdala habituation as a result of cortical lesion?

Cyriel Pennartz asked what is was that limits the stress response. Someone thought that glucocorticoid receptors in the hippocampus might be key, along with projections linking the hippocampus to the hypothalamus.

Tony Grace indicated that in studies of attention deficit hyperactivity disorder, one had to first rule out child abuse, because this can lead to symptoms that resemble ADHD. ADHD may be associated with lack of PFC involvement - functional rather than structural hypoactivity - and failure to suppress inappropriate behaviors. The studies reporting that ADHD treatment is correlated with reduced tendency for substance abuse later in life were mentioned, but then others added that these studies were controversial.

Bita Moghaddam suggested that cortical responses to stressful cognitive tests (e.g., mental arithmetic) can and should be performed in several patient populations, none of whom have yet been tested in this way: drug addicts, abused children, ADHD patients.

Someone asked about the latter part of the first question focusing on changes in limbic and paralimbic regions. In this context, increased monoamine release and sensitization were mentioned as possible alterations in circuitry. Tony Grace talked about his studies with chronic cold stress and changes in locus coeruleus neuron responses. A heightened response like sensitization occurs in these neurons that not only persists, but increases over time.

Anne-Marie Thierry mentioned a study in which rats were isolated for 3 weeks and then housed together after that. This treatment was associated with a long-term increased susceptibility to stress and a decreased turnover of dopamine in the PFC. These two changes were negatively correlated. It was speculated that the increased stress sensitivity might be secondary to having less dopamine and that this would subsequently compromise the ability of the PFC to develop effective coping strategies for dealing with stress.

Weinberger related studies with COMT knockout mice, which exhibit increased PFC DA levels. Compared to wild type mice, these animals are more successful in cognitive tasks when they are stressed.

It was then noted that dopamine in the PFC is not subject to extensive reuptake into dopamine terminals. Instead it is cleared to a certain extent by norepinephrine terminals. As a result, drugs that block norepinephrine uptake also increase levels of dopamine. Due to the sparse dopamine reuptake, it was suggested that PFC dopamine terminals under stressful conditions might be exhausted faster then dopamine terminals in other regions. PFC dopamine terminals would therefore be more dependent on the synthesis of new dopamine. At the least, these ideas highlight the importance of interactions between dopamine and norepinephrine.

Someone suggested that in PTSD a few highly stressful experiences can cause an emergent syndrome that looks like sensitization. However, most people are resistant, even abused children.

The discussion then turned to the new marketing of norepinephrine selective uptake blockers as treatment for ADHD. The main effects of these compounds are thought to be mediated via altered levels of hypothalamic norepinephrine and PFC dopamine. These drugs have low abuse liability.

Charlie Bradberry asked whether there were any concurrent studies looking at the interaction between amphetamine administration and stress. Someone (Weinberger?) replied that there is likely to be an interaction, because some people do better or worse on amphetamine depending on their COMT phenotype and their levels of PFC dopamine. At the least, amphetamine would upset the balance between cortical and subcortical dopamine levels.

The group then considered the third question, beginning with Ahmad Hariri, who wondered whether bipolar disorder might represent a model for affective switching. Julie Fudge added that bipolar disorder involves both a faulty switch and a faulty rheostat. It was asked whether patients with orbitofrontal lesions were depressed. They seem to have a loss of behavioral initiation but not necessarily depression, perhaps depending on the side of the lesion. Instead, orbitofrontal patients seem to have a loss of appropriate behavior.

At this point, the group decided that it might be a mistake to assign too much credit to the PFC for so many functions. Weinberger noted that some investigators consider the caudate nucleus to be the source of problems in ADHD. In this case, the PFC would be operationally normal but disconnected. It was mentioned that in an emotional "Stroop" task, different areas of the PFC are activated in ADHD patients - dorsolateral instead of orbitofrontal areas.

Someone Haber asked whether there was any role for the thalamus, seeing that it gets so many inputs from these areas, including the PFC. Hans Breiter noted that imaging studies indicate that the orbitofrontal gyrus is the single most common area of alteration in psychiatric illness, followed by the amygdala. He went on to indicate that the thalamus has not been as well examined in part because it is hard to dissect out in imaging studies.

A participant urged a return to the question of how to study affective switching. Breiter stated that he did not believe in switching. In his view, mood is more dimensional, less categorical, except when it is abnormal, for example in bipolar disorder. Perhaps medication-induced switching could be scanned, but this had not yet been done.

Weinberger asked why it is that amphetamine is not a better antidepressant, although it does work in the elderly to some extent. Depressed patients who take amphetamine feel somewhat better but they don't get over their condition. A participant suggested that the problem may be due to the fact that amphetamine can directly release catecholamines, which ultimately leads to the offsetting effect of catecholamine depletion.

Another participant asked why chronic selective serotonin reuptake inhibitors are anxiolytic. One respondent felt that it was due their restoration of homeostasis with a slow time course. In other words, slow modulators can cause changes in other monoamines that eventually restore the system to normal. Another person remarked that serotonin transporter knock-out mice are anxious, although this might be a developmental issue. It was also noted that genetic predisposition may affect an individual’s response to SSRIs. In this regard, perhaps obsessive-compulsive disorder, depression and anxiety are not so distinct in their neurobiology. It was also mentioned that SSRIs might not be great antidepressants. Instead, they may be anxiolytic, and then patients improve their mood due to greater ability to create coping strategies. It was pointed out that treatment of OCD with SSRIs can actually take 4-fold higher doses than those used to treat other disorders and sometimes as long as 8-12 weeks of continuous therapy to see an effect. Moreover, combined norepinephrine and serotonin uptake blockers are better antidepressants than SSRIs alone. Current drugs have reduced side effects compared to tricyclics, but they are no more effective as antidepressants. It seems that the best drugs block multiple monoamine uptake carriers.

It was mentioned that cognitive based therapy is as effective as SSRIs and that there is a significant placebo effect in treating depression, indicating that expectancy helps to determine outcome. This prompted the comment that the brain is an inference machine. Money for example has no value by itself, but it produces expectancy and so is a conditioned reward.

Someone suggested that neurogenesis might contribute to the therapeutic effects of antidepressants. In this sense, seizures are known to enhance neurogenesis and so perhaps SSRIs might do the same, i.e., through a toxic rather than a beneficial effect.