Repeated stress induces dendritic spine loss in the rat medial prefrontal cortex

Radley, J.J. et al. (2005). Repeated stress induces dendritic spine loss in the rat medial prefrontal cortex. Cerebral Cortex, 16, 313-320.

The medial prefrontal cortex (mPFC) plays an important role in higher cognitive processes and in the regulation of stress-induced HPA axis activity. This study investigated the effect of stress on dendritic spine density in the mPFC. Rats were restrained for 6 hours daily for 21 days with wire mesh. Following the 21 days of stress, stressed rats weighed less than controls, had a 20% decrease in overall apical dendritic length, a 16% decrease in apical dendritic spine density, and hence an estimated 33% reduction in the total number of axospinous synapses on apical dendrites of pyramidal neurons in the mPFC. These morphological changes may have a significant impact on the functional properties of this region. Clinically, mPFC dysfunction is associated with PTSD and depression. One potential neuroanatomical substrate relevant to these disorders is the mPFC-amygdala circuit. Normally, the mPFC may inhibit amygdala output through its connections on the GABAergic intercalated cells at the border of the lateral and central nuclei of the amygdala. Experimental lesions of the mPFC support this, leading to an enhancement of amygdala-dependent behaviors such as emotionality and fear conditioning. Future studies are needed to investigate the extent to which these morphological changes from chronic stress are reversible.

Neural mechanisms of extinction

Quirk, G.J. & Mueller, D. (2007). Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology Reviews, 1-17.

The simplest form of emotional regulation is extinction, is which conditioned responding to a stimulus decreases when the reinforcer is omitted. Exinction, like any learning process, occurs in 3 phases: acquisition, consolidation, and retrieval. Cannabinoid and opioid receptors appear to be implicated in the acquisition of extinction since anandamide and opioid antagonists impair within-session extinction of fear. Consolidation appears to depend on protein synthesis within the BLA, frequency bursting of the infralimbic region (IL) of the vmPFC shortly after extinction, and general involvement of the hippocampus, especially in tasks such as inhibitory avoidance and contextual fear. Retrieval of extinction memories involves the expression of inhibitory circuitry and is highly context-specific. Inhibition circuitry within the amygdala includes local inhibitory neurons within the BLA and CE, as well as islands of GABAergic neurons between these two sites known as the intercalated (ITC) cells. ITC cells could serve as a site of extinction memory since they inhibit CE output neurons and BLA neurons, acting as an off-switch for the amygdala. ITC cells receive strong projection from the IL mPFC, and IL activity is correlated with the extent of extinction retrieval. In fact, electrical stimulation of IL reduces conditioned fear and strengthens extinction memory. The prelimbic (PL) mPFC, on the other hand, excites fear expression and can augment fear expression via projections to the basal nucleus of the amygdala. Thus, the PFC can fully control overall fear expression. Individuals with PTSD show reduced vmPFC and hippocampal volume and activity, as well as increased amygdala activity. Stress may also impair extinction, since chronic stress is shown to decrease dendritic branching and spine count in hippocampus and mPFC, but increase it in BLA, which could be expected to increase conditioning and impair extinction. Pharmacological adjuncts to current extinction-based exposure therapies may accelerate and strengthen extinction. Among them D-cycloserine, yohimbine, sulpiride, and methylene blue show promise. Administration of glucocorticoids such as cortisol before exposure therapy may also help.

Prolonged Exposure Treatment for PTSD following 9/11

Kazi, A., Freund, B., & Ironson, G. (2008). Prolonged Exposure Treatment for Posttraumatic Stress Disorder following the 9/11 attack with a person who escaped from the Twin Towers. Clinical Case Studies, 7, 100-116.

This article chronicles the progress of one 9/11 survivor through the cognitive-behavioral therapy intervention of prolonged exposure (PE) therapy to address her PTSD and depressive symptomatology. This treatment consists of (1) imaginal exposure, and (2) in vivo exposure. It is designed to elicit emotional processing until the detrimental traumatic memories and avoidances have habituated (desensitized). After 15 sessions this client improved 75% as measured by a composite index. However, there was residual symptomatology 6 months after therapy ended but measures remained sub-clinical. Progress through treatment can be seen as waxing and waning, but trending towards improvement. Still, in this type of therapy clients must be stressed before they are to feel better. With the prevalence of PTSD at 8% in the US population, clinicians are calling more and more for effective treatment regimes. PE may be a promising candidate.

HPA Axis, neuroendocrine factors, and stress

Tsigos, C. & Chrousos, G.P. (2002). Hypothalamic-pituitary-adrenal axis, neuroendocrine factors, and stress. Journal of Psychosomatic Research, 53, 865-871.

Psychology is concerned with the transactions and interactions we have with the world. Stress research examines how we respond to transactions that are stressful. This paper does a great job to illustrate the complexity of our physiological responses to stressors.

CRH/AVP. The paper begins with our physiological response following detection of a stressor. CRH and AVP are secreted into a special portal system and activate neurons of the paraventricular nuclei (PVN) of the hypothalamus, which primarily kicks off activation of the greater hypothalamic-pituitary-adrenal (HPA) axis.

LC/NE system. The locus ceruleus and other noradrenergic cell groups of the medulla and pons, collectively known as the LC/NE system, serve as a global alarm system, using brain epinephrine to execute autonomic and neuroendocrine responses.

The autonomic axis. The ANS provides rapid response to stress, engaging the SNS and withdrawing the PSNS, and enacting cardiovascular, respiratory, gastrointestinal, renal, and endocrine changes.

The HPA Axis. CRH and AVP normally follow certain circadian rhythms, with increasing pulses seen in the early morning hours and decreasing throughout the day. During acute stress, pulsations in this portal system markedly increase, resulting in release of ACTH from the pituitary into the general bloodstream, which finally results in secretion of cortisol and other glucocorticoids from the adrenal cortex. These corticoids involve the whole body in the organism's response to stress and ultimately contribute to the termination of the response via inhibitory feedback.

Other changes. Concomitant with the aforementioned changes, the sympathetic-adrenomedullary system (SAM) influences the body organs, and vagal and sacral parasympathetic responses are also instantiated mediating our gut responses to stress.

The paper also goes into advanced topics including interactions that exist between the HPA axis and the immune system, interactions between the HPA and the gonadal and growth axes, and interactions between the HPA and metabolism. It also discusses pathologies related to the HPA axis. A spectrum of conditions may be associated with increased and prolonged activation of the HPA axis, including melancholic depression, anorexia nervosa, OCD, panic anxiety, excessive exercising, and childhood sexual abuse. Another group of conditions may be associated with hypoactivation of the stress system, including atypical depression, seasonal depression, and chronic fatigue syndrome. Antalarmin, a CRH-R1 antagonist, was also mentioned as being a potentially important drug to combat HPA axis disorders characterized by HPA and LC/NE hyperactivity in the future.

Neural mechanisms of extinction learning and retrieval

Quirk, G.J. & Mueller, D. (2008). Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology, 33, 56-72.

Early in the study of classical conditioning, Pavlov observed spontaneous recovery of responding to an extinguished conditioned stimulus. This witnessed 'uncovering phenomena', brought on by a change in context or stimulus presentation, led to the belief that extinction is not erasure of a previous fear memory but rather the learning of an additional inhibitory memory. This paper reviews what has been learned about extinction learning ever since. Like other types of learning, extinction occurs in three phases: acquisition, consolidation, and retrieval. Acquisition seems to depend on the basolateral amygdala (BLA) and the ventrolateral periaqueductal gray (vlPAG) structures. Consolidation seems to be most dependent on the BLA, where the learning of new memories (requiring protein synthesis) appears to take place. It also seems to rely on the involvement of the prefrontal cortex and the hippocampus. Retrieval requires expression of an inhibitory memory, and as such, during retrieval we see activation of inhibitory networks in the amygdala, cortical control of amygdala inhibition by the IL mPFC, and contextual regulation provided by the hippocampus and mPFC. Thus, like classical conditioning, extinction seems to be distributed across a network of structures, rather than centered in any one particular area.

The paper also raises many fascinating special issues. One challenges the Pavlovian idea that extinction is purely an additional inhibitory memory: recent evidence seems to indicate that extinction leads to reversal of conditioning-induced phosphorylation of CREB, indicating some erasure of the original BLA fear memory!

It also discusses anxiety disorders and PTSD which may be caused by a failure to retrieve an extinction memory generated in extinction-based treatment. Subjects with PTSD show reduced volume and activity in the vmPFC and hippocampus areas, along with increased activity in the amygdala, suggesting inhibitory control and contextual modulation of extinction may be compromised. It also mentions that chronic stress can impede extinction-based therapies, decreasing dendritic branching and spine count in the vmPFC and hippocampus, and increasing dendritic branching and spine count in the BLA, thereby enhancing conditioning effects and impairing extinction. Recently, deficits in fear extinction observed in these human disorders have been combated with pharmacological agents, facilitating extinction of the fear memory with the help of D-cycloserine and impairing fear memory reconsolidation with the B-adrenergic receptor blocker propranolol.

Direct and indirect activation effects on reconsolidation in amygdala

Debiec, J., Doyere, V., Nader, K., LeDoux, J.E. (February 28, 2006). Directly reactivated, but not indirectly reactivated, memories undergo reconsolidation in the amygdala. PNAS, Volume 103, Number 9, 3428-3433.

The first experiment of the paper used second-order fear conditioning (SOFC) to create an associative memory network in rat brain. To do this, a conditioned stimulus (CS1) is paired with an unconditioned stimulus (US), which naturally elicits a response. After pairing, now CS1 elicits the response (such as freezing in fear). This is the first-order conditioning. Now, a second conditioned stimulus (CS2) is paired with CS1, and by association elicits the response transitively. When extinction of CS1 responding does not affect the responding of CS2, CS2 is considered independent of the first-order fear memory. But if CS2 responding decreases with CS1 extinction, then we have an associative chain (CS2 --> CS1 --> US).

The second experiment, building on such a conditioning chain, discovered that extinction of freezing responses to the first-order stimulus (CS1) leads to responding impairments in CS2. Extinction of the second-order stimulus (CS2), does not have any effect on CS1. This builds a case for a hierarchical, uni-directional chain.

The last experiment examined the effect of activation (memory retrieval) on such an associative chain. (In another paper, Nader and LeDoux showed that reactivation of a memory places it in a labile state -- that is, susceptible to disruption -- until again reconsolidated.) Results demonstrated that protein synthesis inhibition after exposure to a single CS1 impairs responses to both CS1 and CS2. But protein synthesis inhibition after exposure to a single CS2, only disrupts CS2 and leaves CS1 freezing intact. Therefore, it is believed that when the first-order association is directly activated, it is placed into a labile state, which may have an impact on dependent associations. However, when the first-order association is only indirectly activated (through an associative chain), it appears that there is not sufficient stimulation to kick off cellular processes which would place it in a labile state, so it remains fixed.

Clinical applications of such research may be in the areas of PTSD, where victims suffer not only from fearful memories, but also from everyday stimuli somehow associated with the initial trauma. This study shows that disrupting associated reactions will only alleviate the sufferer from these quirky stress reactions, while breaking associative chains at the root cause may provide cascading relief.

Facilitation of Extinction of Conditioned Fear by D-Cycloserine

Davis, M., Myers, K.M., Ressler, K.J. Rothbaum, B.O. (2005). Facilitation of Extinction of Conditioned Fear by D-Cycloserine. American Psychological Association, Volume 14, Number 4, 214-219.

Conditioned fear can be suppressed through a process known as extinction, in which repeated exposure to a fearful stimuli minus any aversive effects leads to a gradual reduction in the fear response. Extinction is not thought to be an "unlearning" process, but rather the addition of a supplementary learning process which actively inhibits or suppresses fear responses that are no longer adaptive. Although not much is known about the neural underpinnings of extinction, it is known that the process depends on NMDA receptors within the amygdala which help to consolidate extinction. The compound D-cycloserine binds to the NMDA receptor and improves its efficiency, and administration of D-cycloserine in rats was shown to dose-dependently enhance extinction.

Taking this laboratory research to the clinic, researchers were curious to see if similar effects would be witnessed in human subjects. This article showed that D-cycloserine used alongside exposure therapy for acrophobics (people suffering from a fear of heights) resulted in significantly larger reductions in phobic symptoms and faster improvement as compared with placebo-controlled subjects. While this paints an optimistic picture for the use of D-cycloserine with phobias, it remains to be seen whether the compound will show equally impressive results in improving cognitive behavioral therapies for more complex anxiety disorders such as PTSD.

Behavioral control, the medial prefrontal cortex, and resilience

Maier, S.F., Amat, J., Baratta, M.V., Paul, E., & Watkins, L.R. (2006). Behavioral control, the medial prefrontal cortex, and resilience. Dialogues in Clinical Neuroscience, Vol. 8, No. 4, 397-406.

In a previous article reviewed, "Stressor controllability and learned helplessness: The roles of the dorsal raphe nucleus, serotonin, and corticotrophin-releasing factor", Maier et al showed that the degree of control an organism has over a stressor modulates the impact of this stressor. But how, exactly, controllability enters the equation neurobiologically was undiscussed. This article tackles that issue with some interesting conclusions drawn.

Although the dorsal raphe nucleus (DRN) plays a critical role in learned helplessness, it is unlikely that it is the brain structure responsible for detecting whether or not a stressor is under behavioral control or not. It has neither the processing power nor the appropriate inputs to make such an assessment. However, the ventral medial pre-frontal cortex (vmPFC) is thought to be that structure. Electrical stimulation of this area leads to inhibition of serotonin neurons in the DRN. And inactivating this region eliminates the differential effects of controllability -- that is, both inescapable and escapable shocks produce the same behavioral outcomes. Further, directly activating the vmPFC during inescapable and escapable shocks produces responses in the both groups equivalent to an escapable shock, eluding learned helplessness.

So-called 'immunization effects' are also interesting. Initial experiences with controllable shock appears to attenuate the typical behavioral response to a later exposure to uncontrollable shock. The article shows how the vmPFC becomes associated with the stressor during controllable shock trials and later becomes re-activating again during subsequent uncontrollable trials, thereby inhibiting the DRN.

Furthermore, this 'trained' vmPFC projects to other structures besides the DRN. One of particular importance is the amygdala which is known to play a critical role in classical fear conditioning. The article shows how the pathway from the vmPFC to the central nucleus of the amygdala (CE) can be used to inhibit CE function, thereby retarding fear conditioning. As has been shown, inescapable shock prior to fear conditioning exaggerates fear conditioning. But escapable shock, a stressful event in itself, before fear conditioning is actually shown to reduce the phenomena of fear (in comparison with never-before-shocked animals). The authors remarked, "We know of no other position that would predict, or even explain, how exposure to a highly stressful event could retard the later development of fear." In other words, repeated exposure to aversive stimuli hyper-sensitize the animal to fear. But, repeated exposure to aversive stimuli which the animal can exercise some control over actually helps to desensitize the animal to fear by involving the vmPFC. Still more encouraging, this mechanism of resilience may generalize broadly to quite different situations and circumstances.

The Amygdala

LeDoux, J. (October 2007). The amygdala, Current Biology, Vol. 17, R868-R874.

This is a fantastic overview of a very important brain structure by one of the world's leading experts on the subject. The article covers the amygdala's anatomic organization (its different nuclei), its connectivity (both inputs and outputs), its cellular mechanisms (neurotransmission and neuromodulation), its role in emotional processing, and its implication in a variety of human disorders. The article even covers neuronal processes related to classical conditioning (i.e. fear memory consolidation and reconsolidation), such as changes in synaptic strength (upregulation of post-synaptic receptors by LTP), structural changes in synaptic connectivity (post-synaptic cytoskeletal alterations, presumably dendritic and microtubule changes), and pre-synaptic feedback mechanisms (release of nitric oxide as a messenger).

Affect Regulation

Schore, A.N. (2002). Dysregulation of the Right Brain: A Fundamental Mechanism of Traumatic Attachment and the Psychopathogenesis of Posttraumatic Stress Disorder, Australian and New Zealand Journal of Psychiatry, 36, 9-30.

This article is by Allan Schore, a doctor in UCLA’s department of Psychiatry and an impressively well-published author. Schore’s style is to assimilate interdisciplinary literature into unified theories of brain function. This review article does well to illustrate his approach by integrating the latest thinking from attachment theory, affective neuroscience, developmental stress research, and infant psychiatry into a theory on post-traumatic stress disorder (PTSD). The article drives home several key points:

(1) The individual response to stressful stimuli may or may not be adaptive.

(2) Current evidence shows that the neural circuitry of the stress system is located in the early developing right brain, the hemisphere that is dominant for affect regulation and inhibitory control.

(3) The development of the right brain is highly experience-dependent, and this experience is primarily mediated by the dyadic attachment relationship that develops between caregiver and infant. Simply put, the mother plays a key role in co-regulating the infant’s postnatally developing nervous system, particularly its stress responses. At a very early age, we thereby rely on and learn strategies from our mothers on self-regulation of emotion. As such, if either the mother or the child is improperly psychobiologically attuned to the body-based states of the self or the other (for whatever reason), this can have detrimental effects on the child’s autoregulatory mechanisms which are still under construction.

(4) This point I will offer more parenthetically. Social stressors are far more detrimental than non-social aversive stimuli. As an example, abuse or neglect would be likely to have a much more deleterious effect on the infant brain than assaults from the nonhuman or inanimate, physical environment. From this, I think one can conclude our brains are more specially honed to social cues than to other types of input.

(5) Due to the shape of the human developmental arc, early experiences in life may be particularly important in shaping an individual’s responsiveness later in life. As developmental effects are almost always cumulative, building on the brick and mortar that has already been laid by previous developmental processes, later growth is “limited by the adequacy of already-formed, underlying networks, and therefore maturation is optimal only if the preceding stages were installed optimally”.

(6) Stress effects are also shown to be cumulative. Whereas acute stress produces short-term and reversible deficits, repeated, prolonged, chronic stress can lead to irreversible or only partially reversible enduring effects.

Add it all up and essentially Shore has conceived of a vignette in which the experiences of the maturing infant can establish inefficient coping mechanisms and individual “dissociation” in times of stress, seen behaviorally even many years later in adulthood. The crux of it… “Optimal attachment experiences allow for the emergence of self-awareness, the ability to sense, attend to, and reflect upon the dynamic changes of one’s subjective self states, but traumatic attachments in childhood lead to self-modulation of painful affect by directing attention away from internal emotional states.” More impressive yet are the 246 citations he makes in an 18-page article!

What is personally interesting to me is the take-away that our development – and in this case, our emotional development – is so unambiguously externally mediated, especially by close parental and familial relationships. (We actually see a reasonable case for the intergenerational transmission of regulation and coping strategies.)

But more importantly, we are essentially “propped” or “wired up” to learn via social mechanisms from people with whom we have intimacy. Said another way, certain parts of the brain may be more receptive to reprogramming by social “interfacing” than by other mechanisms, perhaps even in adulthood. This idea is central to Schore’s thinking on the therapist-patient relationship, which he expounds upon in his three-book-set on Affect Regulation.