Biosublime.com/LSD, Birth & TraumaThe Biological Role of the
Endogenous Hallucinogens
Your Subtitle text
REM Sleep Healing Role
R1 REM Sleep Heals Noxious Memories
The main clue to the idea of the healing role of REM sleep is the mystifying character of M’s flashbacks, taking on a life of their own and completing the disappearance or erasure of the noxious memory. This process repeated itself spontaneously soon after awakening in the morning and at certain times of the day without any apparent input from the outside. These were exact encores of the original LSD recall and its somatic effects diminished little by little until their last traces were gone, never to return. The palate pressure was the last sensation to disappear after a few weeks, which was consistent with the area of highest pressure from the obstetrician's fingers. Apparently, a natural spontaneous process exists that expresses some genetically determined function to clear out the baggage of noxious memory, roughly analogous to the healing of a physical wound. Such a process might be expected to follow a regular and continuing schedule to maintain the brain's uncluttered function during the day.
The main suspect would be the sleep stage REM sleep (REMS), a condition that occurs in several mammals and is associated with vivid dreams in humans. Here, the REM state of sleep is examined as involving the same hypothetical mechanism as the RaRN model for erasing noxious memory by the nightly secretion of the endogenous 5-HT1a agonist, DMT+ secreted from the pineal gland. The arguments involve: 1) Other documentation of REMS as a healing state, 2) REMS as a hallucinogenic state, 3) Revisiting the nature of the RaRN memory, as a nightly schedule for opening, releasing and erasing trauma (as defined here), 4) REMS dreams vs NREMS dreams, 5) The pineal gland as a closed delivery system.
1) REMS as a healing state. The conjecture of healing during sleep and REMS in particular is not new. The necessity of REMS has been shown by the disturbing effects on the daily functioning of subjects deprived of the REMS period for two or three days. Carl Jung documented the gradual spontaneous disappearance of (negative) personal issues in sequential dreams long ago (Jung, 1943). Sir Francis Crick (of DNA fame) viewed REM sleep as a state in which memory traces are "forgotten" (Crick and Mitchison, 1983) as a nightly means to re-organize neural memory "nets" by "reverse learning" of unneeded "traces". The Crick-Mitchison theory was built upon and in agreement with earlier work with the same basic view (Newton and Evans, 1965; Gaarder, 1966). Francine Shapiro and Margot Forrest among others have taken the putative trauma-REM connection into practice as a successful therapeutic tool (EMDR) to alleviate trauma and its stressful effects (Shapiro and Forrest, 1997). Shapiro came upon this notion as she relieved uncomfotable mental states by moving her eyes rapidly. Here, the healing connection to REM arrives from a different and independent direction, i.e., the REM-flashback connection proposed here. As a function of REM sleep, the spontaneous and weakening persistence of flashbacks monitoring the progress of noxious memory erasure would take the altered form of repeated sequential dreaming. Testing of this REM-RaRN idea should be accessible with established neurobiological methods to provide the scientific support to answer criticisms of EMDR therapy and provide a pharmacological basis for increasing its efficacy.
2) REMS is Hallucinogenic. As a nightly schedule for opening, releasing and erasing trauma (as defined here), REM sleep has certain attributes that are provocative with respect to its similarity to well-known earmarks of the hallucinogenic state: 1) Like the hallucinogenic state, REMS involves reticular activation, and increased alertness, well documented by the similarity of REM EEG tracings to those of the awake state and dehabituation and awareness of surroundings in hallucinogen inebriation (see this link, 1.8.2.1). This alertness reflects thalamocortical activation seen in a REM-fMRI study (Wherie et al, 2007) and involves brainstem reticular activation with contributions from the orexin system of the hypothalamus, which is subject to 5-HT1a-ligand binding (Muraki et al, 2004). 2) The inability to move, as in REM atonia, is seen on occasion in human and animal subjects on higher doses of both indole and phenethylamine hallucinogens. Like the RaRN -raphe model, muscle atonia is reported to be mediated by the "medullar" raphe nucleus (Hoffman et al, 2007; Brown et al, 2008) and is seen in hallucinogen-induced catalepsy as a form of atonia (Chiu and Mishra, 1980). 3) Similarities are seen in brainstem neurotransmitter sites controlling both REMS and the RaRN model (See "The Brainstem and REM sleep, below) and 4) The distorted imagery and intensity of REMS dreams conjure up the many similar characteristics of hallucinogenic visions, which could easily be considered as the cerebral form of flashbacks. The amygdala is involved in REMS dreams (McNamara in "What are Dreams?"NOVA, PBS). 5) Like the many artistic and literary works inspired by LSD influence, evidence has been gathered to suggest that the REMS is a higher creative state than expected (Medmick in "What Are Dreams?" NOVA PBS). Another clue can be found in the well researched book, "At Day's Close" by A. Roger Ekirch, describing nighttime life before artificial illumination. 14th to 17th century people throughout Europe, retiring early around sunset, would awaken at midnight for an hour or two of notably creative activity before returning to sleep again. REMS, the most likely state previous to awakening, appears with a citation on the altered state of this midnight break (Ekirch, 2005; pp. 322-323). Although saccadic eye movement that defines REMS is not noted in reports of hallucinogen inebriation, it originates from a pons raphe nucleus under the command of the habenula, closely associated with the pineal gland (see Habenular Influence below).
3) The RaRN memory. The nature of hidden memory defined here by the RaRN model is that it lingers after a noxious insult to affect behavior of the recipient indefinitely through unconscious (subcortical) pathways (link 1.4.1). As such, it qualifies as trauma, engrams, memory traces or mnemes, being clearly distinct from the kind of memory that can be recalled willfully. With recall of RaRN memory, one re-lives the sequence of insults; They are not remembered in the abstract as they would be from declarative memory. The absolute requirement for activating and opening this memory for conscious awareness is action of the 5-HT1a agonist of the raphe serotonergic neurons to allow reticular activation and the flow of impulses from the memory storage site into cerebral interpretation. As a pharmacological event, the memory can be opened and closed reversibly for erasure or re-consolidation to produce an updated memory. This is borne out by these results of birth memory retrieval by LSD. Although this particular case of physical skull sensations would likely involve a cerebellar storage site closer to the caudal brainstem and peripheral nervous systems, the documented enervation of the amygdala and the hippocampus by the medial and dorsal raphe nuclei suggests that these could be involved in the same kind of storage model as RaRN, which would contain emotional and psychological input. Insults of this latter kind could be mediated by the DMT+ RaRN mechanism as well. Such speculation is informed by a special arrangement of raphe enervation in one part of the hippocampus (XXX), suggesting long-term memory storage within the hippocampus itself. I
In any case, PTSD is now defined more by the action of DMT+ than by the intensity of the insult. Theoretically, any kind of sensory or cerebral input can be stored as hidden memory when DMT+ happens to be secreted at that time. The effect of DMT+ during a normal birth to open fetal memory sites would be associated with wholesome input from the sensory feedback of its own movements (see Part 2). However, the traumatic nature of the memory in this instance of skull events would be consistent with the argument that REMS concerns negative memory to clear brain areas chronically under the burden of its effects and integrate this memory into cortical declarative memory, now unable to compete for CNS energy and blood within the larger neural volume of the cerebral hemispheres. The stewardship of energy and blood by the brain has been revealed by fMRI, demonstrating the diversion of signal hyperintensity away from unused areas during specific mental tasks.
4) The REMS Dream. Dreams have been the main preoccupation of sleep researchers, who examine the deep ("slow wave") and REM stages of sleep for their characteristic dream imagery. The stages of sleeping have been delineated by the changes in EEG patterns that reveal four stages of "deep" sleep or non-REM sleep (NREMS) and a fifth category, rapid eye movement sleep (REMS). The four consecutive periods of NREMS are stage 1, the shallow onset of sleep, and the remaining three are deep "slow wave" states. Phases 2,3 and 4 are often lumped into one, Phase 2 for "slow wave" experiments. Recent studies on both humans and rats have revealed clear differences in the quality of NREMS and REMS dreams (McNamara et al, 2009; Barrett & McNamara, 2007; Lee & Wilson, 2002; Foster & Wilson, 2006). NREMS dreams are short and compressed, leaving a positive sense and favorable self-image in a subject awakened during this time. On the other hand, REMS dreams are of long duration, are vivid and notable for their negative content, leaving the awakened subject with a lowered self-image carried into the day (Robert & Zadra, 2008), regardless of the fact that REMS dreams are seldom remembered.
That there is more than one kind of dream resolves some conflicting opinions of dream researchers. Just as Jouvet (Jouvet, 1962) rejects the forebrain in favor of the brainstem as the origin of dreams, Solms rejects the brainstem in favor of the forebrain. Possibly, they are both correct if Jovet's dreams are REMS dreams and Solm's dreams emerge from NREMS. The brainstem origin of REMS has been established in the earlier work of Hobson and co-workers (Hobson et al, 1979; McCarley & Hobson, 1975) and is discussed in detail below (see R2 below).
Are REMS dreams always negative and, if so, how does this quality arise? In this experiment LSD brought forth skull sensations that would have been painful, but were felt only as pressure accompanied by sensations of anesthesia and cataleptic sedation. Clearly, the memory was of a traumatic nature and the qualities of pressure, anesthesia and catalepsy were linked most likely in a Pavlovian manner of association that was primitive and chiefly subcortical (see Precedent). This actual re-living of physical insults is not the same as REMS dreams described by researchers or by Jung in 1943. For these workers the emphasis is on the psychological imagery and there's little or no reporting on the physical sensations of the dreamer. On the other hand, pain recall could be part of the mix in the vigorous thrashing around by humans and animals deprived of REMS atonia ( ). The sensations experienced as flashbacks throughout the day by this LSD subject must have involved part of the same nightly process needed for the onset of the REMS state. The implication is that REMS dreams are a cerebral form of flashbacks and would belong to a different category than NREMS dreams not associated with noxious memory storage. Anecdotal evidence related to negative dreams can be found that involve recognition of a symbol known only to the dreamer and by association, replays of hidden memory. Soon after awakening, symbols within the dream theater of otherwise distorted imagery are accurately recognized as part of the dreamer’s unique history (Jung, 1943). On reflection, while the dream itself may seem innocuous, the remembered incident related to the symbol has noxious content. In discussing the pineal gland as the source of DMT+ (below), the point of negativity for REMS dreams is explored further (Habenular influence).
It has been suggested by McNamara that chronic depression may be related to the residue of REMS and its negative character (NOVA, PBS "What Are Dreams?": inside the sleeping mind). The RaRN hypothesis would take this further by postulating that some forms of chronic depression are really PTSD arising from the failure of REMS to complete the erasure and resolution of noxious RaRN memory (see R5 below). REMS dreams originating from hidden memory would reflect the noxious nature of the emerging memory, which is quite consistent with reports on the negative character of REMS reported by others. Thus, while dream imagery certainly requires cerebral function, it is likely that some dreams originate from trauma and, by association, from the RaRN mechanism involving a 5-HT1a agonist to open the memory substrate in the cerebellum for peripheral sensation and/or the hippocampus for emotional content for cortical dream imagery.
5) The Pineal Space. It is ironic that the centerpiece in the development of the RaRN model, so attractive as a mechanistic aspect of REMS and essential for all the foregoing hypotheses, is a molecule (DMT) that has never been measured within any larger biological context (see 1.9 "What is the endogenous hallucinogen?).
Just millimeters away from the brainstem raphe nuclei, its secretions are carried by cerebral spinal fluid (CSF) within a small volume (a norm of just 2 ml total) representing the third ventricle, the cerebral aqueduct, the fourth ventricle and the cisterns and channels that provide intimate contact with brainstem elements and begin the entrance of CSF into the spine (Nolte. 2002). This sequestered and exclusive nature of this secretory path would be advantageous for: 1) The parsimonious delivery of melatonin and DMT+ according to diurnal cycles and 2) For the protection of the pineal-brainstem elements from drugs in the blood supply. Notably, this postulate of pineal function and its sequestered area helps to resolve the conflicting evidence that REM sleep is suppressed, not enhanced, by serotonin re-uptake inhibitors (SSRIs) as well as 5-HT1a agonists (Gillin et al, 1996; Wilson et al, 2005) and MAO inhibitors ( (Wyatt et al, 1969; Vogel et al, 1990). The opposite would be expected with reference to the RaRN model if these drugs had full access to brainstem neurotransmitters and their recetors; These drug effects are somewhat misleading, as they involve several neurotransmitters that mediate both REM-ON and REM-Off sites, as discussed below (R2). Access of these drugs to the brainstem from peripheral systemic injection would require a torturous route via the lesser vestibular/brachial arterial supply and would be filtered in the choroidal elements that produce CSF. Furthermore, it has been shown that a barrier exists between the pineaocytes and the blood compartment (Dominguez & Piezzi, 2007). The onset of the REM state is a complex affair involving cholinergic, GABAergic, noradrenergic and nitric oxide synthease action, all candidates for REM-on and REM-off drug interactions affecting REMS (see below). Further ambiguity arises from the choices offered to systemic drugs by both pre- and post-synaptic receptors of the dorsal raphe (cf. Sorensen et al, 2001; Monti et al, 2001; Monti et al, 2002). Clearly the dorsal raphe has a role in REM sleep (McCarley and Hobson, 1979). The telling point is that direct application to the dorsal raphe nucleus is required to induce REMS by 5-HT1a agonists and inhibit REMS by antagonists as discussed below (R2).
Habenular influence. The second teleological hint is the intimate functional and spacial relationship between the pineal gland and the habenula. The right and left habenular elements hug the pea-sized pineal in the epithalamus and share the habenular commisure with the pineal stalk. The lateral habenula (LHB), distinguishable from the medial habenula (MHb), contains connecting neurons that are inhibitory to dopamine neurons in the ventral tegmental area and substantia nigra, as well as the serotonergic neurons within the medial and dorsal raphe nuclei. Thus,excitation of certain LHB nuclei is inhibitory to the dopamine reward center and, like DMT+ inhibitory to raphe 5-HT neurons (Matsumoto & Hikosaka, 2009; Herkenham and Nauta, 1977; Lecourtier and Kelly, 2007). Unlike the DMT+ - 5HT1a (raphe) inhibition of the RaRN model, the habenular effect is immediate, communicating by direct enervation with the medial and dorsal raphe nuclei (Nolte, 2002). This habenular inhibition has been seen to represent a negative feedback system mediated by negative impulses into the LHB (Matsumoto & Hikosaka, 2009). Another raphe nucleus within the pons reticular formation, the raphe interpositus, controls saccadic eye movements, slowing it down under habenular influence ( ). Since the habenular activation follows the diurnal cycle receive impulses of negative quality they mediate negative feedback for movement and positive reward. Both the pineal and the habenula receive sympathetic impulses from the suprachiasmatic nucleus (SCN), which relays photonic information from the eye for the pineal's nightly secretion of melatonin. The location of SCN within the lateral hypothalamus further suggests how negative impulses originating from external perceptions or cerebral thought processes can activate PTSD via the RaRN model for hidden memory.
Once again, the well-known anecdotes of traumatized war veterans enter the discussion. In one, the backfire of an automobile exhaust sends a Vietnam veteran into a physiological fugue beyond his or her control, including brief unconsciousness, tachicardia and high blood pressure. In a recent TV news account, a young soldier returning from Iraq has dinner with some of his stateside platoon mates to share war stories. A day or two later he wakes up in a jail cell, having been arrested for kicking in a neighbor's door and destroying the furniture. His having no (declarative) memory of this destructive behavior is evidence the unconscious power of perceptual input having siblings within hidden memory. Cerebral cortical areas and the limbic system communicate with the dorsal diencephalic communication system (habenula, pineal and connections) through the hypothalamus, containing the suprachiasmatic nucleus (light/dark) and associated elements for interpreting descending impulses.
As already discussed in Part 1, Twilight Sleep, the RaRN model is bidirectional, where recall of noxious memory can further traumatize as long as the memory substrate, now opened by the 5-HT1a agonist, remains receptive to internal and external events. The question is whether trauma gives rise to dreams. During REM sleep the RaRN model might not operate to reconsolidate the memory; it might only be involved in releasing noxious memory for stepwise erasure and lose its bidirectional property. This expectation is based on the fact that M's flashbacks returned as a stepwise depletion of the memory, once the first stimulation of recall by LSD took place. The subject was not overwhelmed by the intensity of the trauma's recall. As a form of healing, flashbacks would signal trauma erasure in a manner similar to the flashbacks M experienced after the first LSD recall, diminishing in a more gentle stepwise manner until the trauma completely disappeared. If REM dreams were actually cortical forms of flashbacks, certain REM dreams would repeat themselves, as bad dreams often do, until the underlying issue disappeared. As a point of nomenclature, dreams that would manifest as trauma recall would be trauma dreams, not traumatic dreams, since the latter would seem to be unlikely if REM sleep can complete its healing process.
The main clue to the idea of the healing role of REM sleep is the mystifying character of M’s flashbacks, taking on a life of their own and completing the disappearance or erasure of the noxious memory. This process repeated itself spontaneously soon after awakening in the morning and at certain times of the day without any apparent input from the outside. These were exact encores of the original LSD recall and its somatic effects diminished little by little until their last traces were gone, never to return. The palate pressure was the last sensation to disappear after a few weeks, which was consistent with the area of highest pressure from the obstetrician's fingers. Apparently, a natural spontaneous process exists that expresses some genetically determined function to clear out the baggage of noxious memory, roughly analogous to the healing of a physical wound. Such a process might be expected to follow a regular and continuing schedule to maintain the brain's uncluttered function during the day.
The main suspect would be the sleep stage REM sleep (REMS), a condition that occurs in several mammals and is associated with vivid dreams in humans. Here, the REM state of sleep is examined as involving the same hypothetical mechanism as the RaRN model for erasing noxious memory by the nightly secretion of the endogenous 5-HT1a agonist, DMT+ secreted from the pineal gland. The arguments involve: 1) Other documentation of REMS as a healing state, 2) REMS as a hallucinogenic state, 3) Revisiting the nature of the RaRN memory, as a nightly schedule for opening, releasing and erasing trauma (as defined here), 4) REMS dreams vs NREMS dreams, 5) The pineal gland as a closed delivery system.
1) REMS as a healing state. The conjecture of healing during sleep and REMS in particular is not new. The necessity of REMS has been shown by the disturbing effects on the daily functioning of subjects deprived of the REMS period for two or three days. Carl Jung documented the gradual spontaneous disappearance of (negative) personal issues in sequential dreams long ago (Jung, 1943). Sir Francis Crick (of DNA fame) viewed REM sleep as a state in which memory traces are "forgotten" (Crick and Mitchison, 1983) as a nightly means to re-organize neural memory "nets" by "reverse learning" of unneeded "traces". The Crick-Mitchison theory was built upon and in agreement with earlier work with the same basic view (Newton and Evans, 1965; Gaarder, 1966). Francine Shapiro and Margot Forrest among others have taken the putative trauma-REM connection into practice as a successful therapeutic tool (EMDR) to alleviate trauma and its stressful effects (Shapiro and Forrest, 1997). Shapiro came upon this notion as she relieved uncomfotable mental states by moving her eyes rapidly. Here, the healing connection to REM arrives from a different and independent direction, i.e., the REM-flashback connection proposed here. As a function of REM sleep, the spontaneous and weakening persistence of flashbacks monitoring the progress of noxious memory erasure would take the altered form of repeated sequential dreaming. Testing of this REM-RaRN idea should be accessible with established neurobiological methods to provide the scientific support to answer criticisms of EMDR therapy and provide a pharmacological basis for increasing its efficacy.
2) REMS is Hallucinogenic. As a nightly schedule for opening, releasing and erasing trauma (as defined here), REM sleep has certain attributes that are provocative with respect to its similarity to well-known earmarks of the hallucinogenic state: 1) Like the hallucinogenic state, REMS involves reticular activation, and increased alertness, well documented by the similarity of REM EEG tracings to those of the awake state and dehabituation and awareness of surroundings in hallucinogen inebriation (see this link, 1.8.2.1). This alertness reflects thalamocortical activation seen in a REM-fMRI study (Wherie et al, 2007) and involves brainstem reticular activation with contributions from the orexin system of the hypothalamus, which is subject to 5-HT1a-ligand binding (Muraki et al, 2004). 2) The inability to move, as in REM atonia, is seen on occasion in human and animal subjects on higher doses of both indole and phenethylamine hallucinogens. Like the RaRN -raphe model, muscle atonia is reported to be mediated by the "medullar" raphe nucleus (Hoffman et al, 2007; Brown et al, 2008) and is seen in hallucinogen-induced catalepsy as a form of atonia (Chiu and Mishra, 1980). 3) Similarities are seen in brainstem neurotransmitter sites controlling both REMS and the RaRN model (See "The Brainstem and REM sleep, below) and 4) The distorted imagery and intensity of REMS dreams conjure up the many similar characteristics of hallucinogenic visions, which could easily be considered as the cerebral form of flashbacks. The amygdala is involved in REMS dreams (McNamara in "What are Dreams?"NOVA, PBS). 5) Like the many artistic and literary works inspired by LSD influence, evidence has been gathered to suggest that the REMS is a higher creative state than expected (Medmick in "What Are Dreams?" NOVA PBS). Another clue can be found in the well researched book, "At Day's Close" by A. Roger Ekirch, describing nighttime life before artificial illumination. 14th to 17th century people throughout Europe, retiring early around sunset, would awaken at midnight for an hour or two of notably creative activity before returning to sleep again. REMS, the most likely state previous to awakening, appears with a citation on the altered state of this midnight break (Ekirch, 2005; pp. 322-323). Although saccadic eye movement that defines REMS is not noted in reports of hallucinogen inebriation, it originates from a pons raphe nucleus under the command of the habenula, closely associated with the pineal gland (see Habenular Influence below).
3) The RaRN memory. The nature of hidden memory defined here by the RaRN model is that it lingers after a noxious insult to affect behavior of the recipient indefinitely through unconscious (subcortical) pathways (link 1.4.1). As such, it qualifies as trauma, engrams, memory traces or mnemes, being clearly distinct from the kind of memory that can be recalled willfully. With recall of RaRN memory, one re-lives the sequence of insults; They are not remembered in the abstract as they would be from declarative memory. The absolute requirement for activating and opening this memory for conscious awareness is action of the 5-HT1a agonist of the raphe serotonergic neurons to allow reticular activation and the flow of impulses from the memory storage site into cerebral interpretation. As a pharmacological event, the memory can be opened and closed reversibly for erasure or re-consolidation to produce an updated memory. This is borne out by these results of birth memory retrieval by LSD. Although this particular case of physical skull sensations would likely involve a cerebellar storage site closer to the caudal brainstem and peripheral nervous systems, the documented enervation of the amygdala and the hippocampus by the medial and dorsal raphe nuclei suggests that these could be involved in the same kind of storage model as RaRN, which would contain emotional and psychological input. Insults of this latter kind could be mediated by the DMT+ RaRN mechanism as well. Such speculation is informed by a special arrangement of raphe enervation in one part of the hippocampus (XXX), suggesting long-term memory storage within the hippocampus itself. I
In any case, PTSD is now defined more by the action of DMT+ than by the intensity of the insult. Theoretically, any kind of sensory or cerebral input can be stored as hidden memory when DMT+ happens to be secreted at that time. The effect of DMT+ during a normal birth to open fetal memory sites would be associated with wholesome input from the sensory feedback of its own movements (see Part 2). However, the traumatic nature of the memory in this instance of skull events would be consistent with the argument that REMS concerns negative memory to clear brain areas chronically under the burden of its effects and integrate this memory into cortical declarative memory, now unable to compete for CNS energy and blood within the larger neural volume of the cerebral hemispheres. The stewardship of energy and blood by the brain has been revealed by fMRI, demonstrating the diversion of signal hyperintensity away from unused areas during specific mental tasks.
4) The REMS Dream. Dreams have been the main preoccupation of sleep researchers, who examine the deep ("slow wave") and REM stages of sleep for their characteristic dream imagery. The stages of sleeping have been delineated by the changes in EEG patterns that reveal four stages of "deep" sleep or non-REM sleep (NREMS) and a fifth category, rapid eye movement sleep (REMS). The four consecutive periods of NREMS are stage 1, the shallow onset of sleep, and the remaining three are deep "slow wave" states. Phases 2,3 and 4 are often lumped into one, Phase 2 for "slow wave" experiments. Recent studies on both humans and rats have revealed clear differences in the quality of NREMS and REMS dreams (McNamara et al, 2009; Barrett & McNamara, 2007; Lee & Wilson, 2002; Foster & Wilson, 2006). NREMS dreams are short and compressed, leaving a positive sense and favorable self-image in a subject awakened during this time. On the other hand, REMS dreams are of long duration, are vivid and notable for their negative content, leaving the awakened subject with a lowered self-image carried into the day (Robert & Zadra, 2008), regardless of the fact that REMS dreams are seldom remembered.
That there is more than one kind of dream resolves some conflicting opinions of dream researchers. Just as Jouvet (Jouvet, 1962) rejects the forebrain in favor of the brainstem as the origin of dreams, Solms rejects the brainstem in favor of the forebrain. Possibly, they are both correct if Jovet's dreams are REMS dreams and Solm's dreams emerge from NREMS. The brainstem origin of REMS has been established in the earlier work of Hobson and co-workers (Hobson et al, 1979; McCarley & Hobson, 1975) and is discussed in detail below (see R2 below).
Are REMS dreams always negative and, if so, how does this quality arise? In this experiment LSD brought forth skull sensations that would have been painful, but were felt only as pressure accompanied by sensations of anesthesia and cataleptic sedation. Clearly, the memory was of a traumatic nature and the qualities of pressure, anesthesia and catalepsy were linked most likely in a Pavlovian manner of association that was primitive and chiefly subcortical (see Precedent). This actual re-living of physical insults is not the same as REMS dreams described by researchers or by Jung in 1943. For these workers the emphasis is on the psychological imagery and there's little or no reporting on the physical sensations of the dreamer. On the other hand, pain recall could be part of the mix in the vigorous thrashing around by humans and animals deprived of REMS atonia ( ). The sensations experienced as flashbacks throughout the day by this LSD subject must have involved part of the same nightly process needed for the onset of the REMS state. The implication is that REMS dreams are a cerebral form of flashbacks and would belong to a different category than NREMS dreams not associated with noxious memory storage. Anecdotal evidence related to negative dreams can be found that involve recognition of a symbol known only to the dreamer and by association, replays of hidden memory. Soon after awakening, symbols within the dream theater of otherwise distorted imagery are accurately recognized as part of the dreamer’s unique history (Jung, 1943). On reflection, while the dream itself may seem innocuous, the remembered incident related to the symbol has noxious content. In discussing the pineal gland as the source of DMT+ (below), the point of negativity for REMS dreams is explored further (Habenular influence).
It has been suggested by McNamara that chronic depression may be related to the residue of REMS and its negative character (NOVA, PBS "What Are Dreams?": inside the sleeping mind). The RaRN hypothesis would take this further by postulating that some forms of chronic depression are really PTSD arising from the failure of REMS to complete the erasure and resolution of noxious RaRN memory (see R5 below). REMS dreams originating from hidden memory would reflect the noxious nature of the emerging memory, which is quite consistent with reports on the negative character of REMS reported by others. Thus, while dream imagery certainly requires cerebral function, it is likely that some dreams originate from trauma and, by association, from the RaRN mechanism involving a 5-HT1a agonist to open the memory substrate in the cerebellum for peripheral sensation and/or the hippocampus for emotional content for cortical dream imagery.
5) The Pineal Space. It is ironic that the centerpiece in the development of the RaRN model, so attractive as a mechanistic aspect of REMS and essential for all the foregoing hypotheses, is a molecule (DMT) that has never been measured within any larger biological context (see 1.9 "What is the endogenous hallucinogen?).
Just millimeters away from the brainstem raphe nuclei, its secretions are carried by cerebral spinal fluid (CSF) within a small volume (a norm of just 2 ml total) representing the third ventricle, the cerebral aqueduct, the fourth ventricle and the cisterns and channels that provide intimate contact with brainstem elements and begin the entrance of CSF into the spine (Nolte. 2002). This sequestered and exclusive nature of this secretory path would be advantageous for: 1) The parsimonious delivery of melatonin and DMT+ according to diurnal cycles and 2) For the protection of the pineal-brainstem elements from drugs in the blood supply. Notably, this postulate of pineal function and its sequestered area helps to resolve the conflicting evidence that REM sleep is suppressed, not enhanced, by serotonin re-uptake inhibitors (SSRIs) as well as 5-HT1a agonists (Gillin et al, 1996; Wilson et al, 2005) and MAO inhibitors ( (Wyatt et al, 1969; Vogel et al, 1990). The opposite would be expected with reference to the RaRN model if these drugs had full access to brainstem neurotransmitters and their recetors; These drug effects are somewhat misleading, as they involve several neurotransmitters that mediate both REM-ON and REM-Off sites, as discussed below (R2). Access of these drugs to the brainstem from peripheral systemic injection would require a torturous route via the lesser vestibular/brachial arterial supply and would be filtered in the choroidal elements that produce CSF. Furthermore, it has been shown that a barrier exists between the pineaocytes and the blood compartment (Dominguez & Piezzi, 2007). The onset of the REM state is a complex affair involving cholinergic, GABAergic, noradrenergic and nitric oxide synthease action, all candidates for REM-on and REM-off drug interactions affecting REMS (see below). Further ambiguity arises from the choices offered to systemic drugs by both pre- and post-synaptic receptors of the dorsal raphe (cf. Sorensen et al, 2001; Monti et al, 2001; Monti et al, 2002). Clearly the dorsal raphe has a role in REM sleep (McCarley and Hobson, 1979). The telling point is that direct application to the dorsal raphe nucleus is required to induce REMS by 5-HT1a agonists and inhibit REMS by antagonists as discussed below (R2).
Habenular influence. The second teleological hint is the intimate functional and spacial relationship between the pineal gland and the habenula. The right and left habenular elements hug the pea-sized pineal in the epithalamus and share the habenular commisure with the pineal stalk. The lateral habenula (LHB), distinguishable from the medial habenula (MHb), contains connecting neurons that are inhibitory to dopamine neurons in the ventral tegmental area and substantia nigra, as well as the serotonergic neurons within the medial and dorsal raphe nuclei. Thus,excitation of certain LHB nuclei is inhibitory to the dopamine reward center and, like DMT+ inhibitory to raphe 5-HT neurons (Matsumoto & Hikosaka, 2009; Herkenham and Nauta, 1977; Lecourtier and Kelly, 2007). Unlike the DMT+ - 5HT1a (raphe) inhibition of the RaRN model, the habenular effect is immediate, communicating by direct enervation with the medial and dorsal raphe nuclei (Nolte, 2002). This habenular inhibition has been seen to represent a negative feedback system mediated by negative impulses into the LHB (Matsumoto & Hikosaka, 2009). Another raphe nucleus within the pons reticular formation, the raphe interpositus, controls saccadic eye movements, slowing it down under habenular influence ( ). Since the habenular activation follows the diurnal cycle receive impulses of negative quality they mediate negative feedback for movement and positive reward. Both the pineal and the habenula receive sympathetic impulses from the suprachiasmatic nucleus (SCN), which relays photonic information from the eye for the pineal's nightly secretion of melatonin. The location of SCN within the lateral hypothalamus further suggests how negative impulses originating from external perceptions or cerebral thought processes can activate PTSD via the RaRN model for hidden memory.
Once again, the well-known anecdotes of traumatized war veterans enter the discussion. In one, the backfire of an automobile exhaust sends a Vietnam veteran into a physiological fugue beyond his or her control, including brief unconsciousness, tachicardia and high blood pressure. In a recent TV news account, a young soldier returning from Iraq has dinner with some of his stateside platoon mates to share war stories. A day or two later he wakes up in a jail cell, having been arrested for kicking in a neighbor's door and destroying the furniture. His having no (declarative) memory of this destructive behavior is evidence the unconscious power of perceptual input having siblings within hidden memory. Cerebral cortical areas and the limbic system communicate with the dorsal diencephalic communication system (habenula, pineal and connections) through the hypothalamus, containing the suprachiasmatic nucleus (light/dark) and associated elements for interpreting descending impulses.
As already discussed in Part 1, Twilight Sleep, the RaRN model is bidirectional, where recall of noxious memory can further traumatize as long as the memory substrate, now opened by the 5-HT1a agonist, remains receptive to internal and external events. The question is whether trauma gives rise to dreams. During REM sleep the RaRN model might not operate to reconsolidate the memory; it might only be involved in releasing noxious memory for stepwise erasure and lose its bidirectional property. This expectation is based on the fact that M's flashbacks returned as a stepwise depletion of the memory, once the first stimulation of recall by LSD took place. The subject was not overwhelmed by the intensity of the trauma's recall. As a form of healing, flashbacks would signal trauma erasure in a manner similar to the flashbacks M experienced after the first LSD recall, diminishing in a more gentle stepwise manner until the trauma completely disappeared. If REM dreams were actually cortical forms of flashbacks, certain REM dreams would repeat themselves, as bad dreams often do, until the underlying issue disappeared. As a point of nomenclature, dreams that would manifest as trauma recall would be trauma dreams, not traumatic dreams, since the latter would seem to be unlikely if REM sleep can complete its healing process.
R2 The Brain Stem and REM Sleep.
Earlier studies on REM neurophysiology pointed to its origin in the rostral or pons area of the brainstem (Jouvet 1962, McCarley & Hobson 1975). As already mentioned, application of agonists and antagonists to the dorsal raphe 5-HT1a receptor seems to be well established in initiating and stopping REM sleep (see below). The dorsal raphe nucleus is itself comprised of six to eight sub-nuclei and is the source of the great majority (ca. 90%) of serotonergic projections throughout the brain. Many attributes of REM sleep, e.g., changes in heart rate, blood pressure and body temperature are reproduced during non-REM sleep (NRM) with the application of raphe 5-HT1a agonist and reversed by 1a antagonists (Brown et al, 2007; Hoffman et al, 2008). NREM + 1a agonist = REM. Other neurotransmitter systems within the brainstem affect REM sleep by interacting with “REM-on” and REM-off neurons acting in reciprocal fashion. REM-on neurons are those that fire during REM sleep and REM-off neurons, e.g., the raphe 5-HT1a neurons, must remain inactive at the same time. The primary requirement for activating REM sleep is the cessation of firing of REM-OFF neurons (Pal and Mallick, 2007). The different locations of REM-on and REM-off neurons are shown in the table below:
TABLE LEGEND:
The headings, "REM-ON" and "REM-OFF" are the starting states before drug application.
ovlPAG ventrolateral periaqueductal gray (Sapin et al, 2009)
dDpMe dorsal part of the deep mesencephalic reticular nucleus immediately ventral to vlPAG (Sapin et al, 2009)
Med-RN medullary reticular nuclei known to generate muscle atonia during REM (Hoffman et al, 2007; Brown et al, 2008)
SLD is non-GABA
Pons Oralis A reticular nucleus in the pons region of the brainstem (Ming-Chu et al, 1999).
RPO nucleus reticularis pontis oralis (Sanford, et al, 2003)
RPC nucleus reticularis pontis caudalis (Sanford, et al, 2003)
Ld/pp/t Laterodorsal/pedunculopontine tegmentum (Pal and Mallick 2007)
Locus ceruleus (Pal and Mallick 2007)
A brain buster: The insertion of "Medllar raphe" as a REM-ON site in the table was made by this author and may be, as the French say, gauche. The placement as REM-on refers to the stopping of REM sleep in piglets with the application of the 5-HT1a agonist, 8-OH DPAT ,to the medullar raphe (Brown et al, 2008). This result seems to contradict a similar experiment by these workers of agonist action on this same medullar nucleus to produce muscular paralysis or atonia. Since atonia is a classic sign of REM sleep, the implication is that the REM state during atonia is being shut down by the same agonist simultaneously. Also, the validity of the Brown, et al observation demands that the dorsal and "medullar" raphe nuclei have opposite functions: inhibiting the serotonergic neurons of the dorsal and "medullar" raphe turns REM on and off, respectively. Consequently, either nightly secretion of the agonist is in a state of conflict or there's a timed protocol. Paradoxes abound, as seen below.
R3 The Opposition: NERUOPHYSIOLOGY
1) Some studies have eliminated the raphe nuclei altogether as mediating the REM state on the basis of retrograde dye transfer to trace the origins of 5-HT neurons in the brainstem (Rodrigo-Argulo, 2000). These serotonergic fibers originate from a different source in a location other than the brainstem.
2) In possible consistency with the Rodrigo-Argulo results, attention is called to the fact that 5-HT1a receptors that do not originate from 5-HT neurons share space within raphe nuclei and. too, are found elsewhere (Hoffman et al, 2007; Brown et al, 2008 see above table legend).
3) Shutting off 5-HT neurons in the medullar raphe of the piglets by the 5-HT1a agonist eliminates REM sleep. The specific binding of agonists and antagonists even in the presence of high 5-HT levels expected in the cooling stress for the animal would support this observation. However, this observation is questionable, as explained above.
Thus, the opening of brainstem memory substrates by a powerful 5-HT1a receptor agonist is a fact (M's LSD recall), but the involvement of raphe nuclei per se is weakened by the discovery of this same 1a receptor elsewhere. However, REM sleep enhancement by the agonist applied directly to the dorsal raphe nucleus encourages the RaRN model as a means to open the memory. There is one more problem.
Another area of conflict is the difference between the hypothesis of the RaRN model and human experimentation on REM sleep. It is proposed here that both DMT trauma recall and REM sleep, as well as the approach to PTSD therapy, begin with the suppression of raphe 5-HT neuron firing by agonist action at the 5-HT1a receptor. However, as mentioned above, REM sleep is suppressed with systemic administration of several 1a agonists of different molecular structures. Consistent with this is the suppression of REM sleep by systemic administration of monoamine oxidase inhibitors (MAOIs), which would be expected to increase the level of monoamines within the blood supply to the brainstem (Wyatt et al, 1969; Vogel et al, 1990). On the other hand, direct application of 1a agonists to the dorsal raphe nucleus promote REM sleep and antagonists suppress this period in animal studies.
A plausable way out of this dilemma is to call forth the argument that the onset of both REM sleep and flashbacks involves a highly localized system of DMT delivery to the brainstem raphe nuclei, which is largely inaccessible to the effects of drugs within the circulatory system. The plausibility of this argument lies in the proximity of the pineal source of DMT to the brainstem (see 1.9 "What is the endogenous hallucinogen?"). Secretion of DMT or DMT+ and their fast delivery to the brainstem would shorten the time of exposure to MAO(a). However, this argument is seriously opposed by the effective treatment of PTSD by systemic administration of serotonin congeners and, notably, MDMA as discussed in the page of this site, "Raphe --- Trauma Therapy" (PTSD is proposed to be accessible therapeutically via the RaRN model). The most likely answer to this is that systemic administration produces drug levels high enough to survive the effects of MAO and overcome the circulatory barriers between the DMT source and the brainstem. The spontaneous secretion of the 5-HT1a agonist for inducing the experienced flashbacks or REM sleep would be impossible without a rapid system for its delivery to the brainstem. This issue will be encountered again in the rebuttal to the claims of Vertes and Marshall (below). Yet, the effects of astonishingly small amounts of (systemic) LSD remains to be explained. This might be answered by the proposal that systemic access of drugs to brainstem functional neurons depends on their relative resistance to MAOa (for the indoles) and MAOb (for the phenethylamines). A comparison of these drugs as to their MAO resistance and their inebriation time would be illuminating. The lifetime of systemic DMT effects are short-lived (Strassman and Qualis, 1994), but is lengthened greatly by MAO inhibitors (Schultes et al, 2001). As the LSD effect averages around nine hours, it's likely that its access and that of other drugs to the raphe nuclei depends mainly upon resistance to MAO oxidation.
Earlier studies on REM neurophysiology pointed to its origin in the rostral or pons area of the brainstem (Jouvet 1962, McCarley & Hobson 1975). As already mentioned, application of agonists and antagonists to the dorsal raphe 5-HT1a receptor seems to be well established in initiating and stopping REM sleep (see below). The dorsal raphe nucleus is itself comprised of six to eight sub-nuclei and is the source of the great majority (ca. 90%) of serotonergic projections throughout the brain. Many attributes of REM sleep, e.g., changes in heart rate, blood pressure and body temperature are reproduced during non-REM sleep (NRM) with the application of raphe 5-HT1a agonist and reversed by 1a antagonists (Brown et al, 2007; Hoffman et al, 2008). NREM + 1a agonist = REM. Other neurotransmitter systems within the brainstem affect REM sleep by interacting with “REM-on” and REM-off neurons acting in reciprocal fashion. REM-on neurons are those that fire during REM sleep and REM-off neurons, e.g., the raphe 5-HT1a neurons, must remain inactive at the same time. The primary requirement for activating REM sleep is the cessation of firing of REM-OFF neurons (Pal and Mallick, 2007). The different locations of REM-on and REM-off neurons are shown in the table below:
TABLE LEGEND:
The headings, "REM-ON" and "REM-OFF" are the starting states before drug application.
ovlPAG ventrolateral periaqueductal gray (Sapin et al, 2009)
dDpMe dorsal part of the deep mesencephalic reticular nucleus immediately ventral to vlPAG (Sapin et al, 2009)
Med-RN medullary reticular nuclei known to generate muscle atonia during REM (Hoffman et al, 2007; Brown et al, 2008)
SLD is non-GABA
Pons Oralis A reticular nucleus in the pons region of the brainstem (Ming-Chu et al, 1999).
RPO nucleus reticularis pontis oralis (Sanford, et al, 2003)
RPC nucleus reticularis pontis caudalis (Sanford, et al, 2003)
Ld/pp/t Laterodorsal/pedunculopontine tegmentum (Pal and Mallick 2007)
Locus ceruleus (Pal and Mallick 2007)
A brain buster: The insertion of "Medllar raphe" as a REM-ON site in the table was made by this author and may be, as the French say, gauche. The placement as REM-on refers to the stopping of REM sleep in piglets with the application of the 5-HT1a agonist, 8-OH DPAT ,to the medullar raphe (Brown et al, 2008). This result seems to contradict a similar experiment by these workers of agonist action on this same medullar nucleus to produce muscular paralysis or atonia. Since atonia is a classic sign of REM sleep, the implication is that the REM state during atonia is being shut down by the same agonist simultaneously. Also, the validity of the Brown, et al observation demands that the dorsal and "medullar" raphe nuclei have opposite functions: inhibiting the serotonergic neurons of the dorsal and "medullar" raphe turns REM on and off, respectively. Consequently, either nightly secretion of the agonist is in a state of conflict or there's a timed protocol. Paradoxes abound, as seen below.
R3 The Opposition: NERUOPHYSIOLOGY
1) Some studies have eliminated the raphe nuclei altogether as mediating the REM state on the basis of retrograde dye transfer to trace the origins of 5-HT neurons in the brainstem (Rodrigo-Argulo, 2000). These serotonergic fibers originate from a different source in a location other than the brainstem.
2) In possible consistency with the Rodrigo-Argulo results, attention is called to the fact that 5-HT1a receptors that do not originate from 5-HT neurons share space within raphe nuclei and. too, are found elsewhere (Hoffman et al, 2007; Brown et al, 2008 see above table legend).
3) Shutting off 5-HT neurons in the medullar raphe of the piglets by the 5-HT1a agonist eliminates REM sleep. The specific binding of agonists and antagonists even in the presence of high 5-HT levels expected in the cooling stress for the animal would support this observation. However, this observation is questionable, as explained above.
Thus, the opening of brainstem memory substrates by a powerful 5-HT1a receptor agonist is a fact (M's LSD recall), but the involvement of raphe nuclei per se is weakened by the discovery of this same 1a receptor elsewhere. However, REM sleep enhancement by the agonist applied directly to the dorsal raphe nucleus encourages the RaRN model as a means to open the memory. There is one more problem.
Another area of conflict is the difference between the hypothesis of the RaRN model and human experimentation on REM sleep. It is proposed here that both DMT trauma recall and REM sleep, as well as the approach to PTSD therapy, begin with the suppression of raphe 5-HT neuron firing by agonist action at the 5-HT1a receptor. However, as mentioned above, REM sleep is suppressed with systemic administration of several 1a agonists of different molecular structures. Consistent with this is the suppression of REM sleep by systemic administration of monoamine oxidase inhibitors (MAOIs), which would be expected to increase the level of monoamines within the blood supply to the brainstem (Wyatt et al, 1969; Vogel et al, 1990). On the other hand, direct application of 1a agonists to the dorsal raphe nucleus promote REM sleep and antagonists suppress this period in animal studies.
A plausable way out of this dilemma is to call forth the argument that the onset of both REM sleep and flashbacks involves a highly localized system of DMT delivery to the brainstem raphe nuclei, which is largely inaccessible to the effects of drugs within the circulatory system. The plausibility of this argument lies in the proximity of the pineal source of DMT to the brainstem (see 1.9 "What is the endogenous hallucinogen?"). Secretion of DMT or DMT+ and their fast delivery to the brainstem would shorten the time of exposure to MAO(a). However, this argument is seriously opposed by the effective treatment of PTSD by systemic administration of serotonin congeners and, notably, MDMA as discussed in the page of this site, "Raphe --- Trauma Therapy" (PTSD is proposed to be accessible therapeutically via the RaRN model). The most likely answer to this is that systemic administration produces drug levels high enough to survive the effects of MAO and overcome the circulatory barriers between the DMT source and the brainstem. The spontaneous secretion of the 5-HT1a agonist for inducing the experienced flashbacks or REM sleep would be impossible without a rapid system for its delivery to the brainstem. This issue will be encountered again in the rebuttal to the claims of Vertes and Marshall (below). Yet, the effects of astonishingly small amounts of (systemic) LSD remains to be explained. This might be answered by the proposal that systemic access of drugs to brainstem functional neurons depends on their relative resistance to MAOa (for the indoles) and MAOb (for the phenethylamines). A comparison of these drugs as to their MAO resistance and their inebriation time would be illuminating. The lifetime of systemic DMT effects are short-lived (Strassman and Qualis, 1994), but is lengthened greatly by MAO inhibitors (Schultes et al, 2001). As the LSD effect averages around nine hours, it's likely that its access and that of other drugs to the raphe nuclei depends mainly upon resistance to MAO oxidation.
On the whole, the functions of the raphe activity appear to be inhibitory and protective to the organism’s homeostasis by preventing sensory overload and untoward activation of specialized brain areas. The suppression of their 5-HT neurons provides a selective dis-inhibition of these specialized areas needed at a particular time. Three properties of REM dreaming associated with protection are: 1) Atonia during the REM state would protect the dreamer from harm by violent movement, which testifies to the high intensity possible in REM dreams, 2) After dreaming, there is a mechanism that militates against recall of the dream that is consistent with the re-closing of the memory substrate as the raphe restore normal activity and 3) Recurring dreams of noxious events imply that, like flashbacks, the release of traumatic material is metered in a sequential, stepwise manner each REM cycle to avoid the release of stored impulses too noxious to bear all at once. This speculation is visited again later in a discussion of the RaRN model’s suitability as a means for re-consolidating noxious flashbacks. This stepwise activation of trauma as recurring (REM) dreams of a particular noxious insult would prevent the reconsolidation of the total insult and provide for sequential erasure of the memory as seen in M’s flashbacks.
These conjectures give rise to two considerations, one about the role of REM dreaming and the other, the relation between REM sleep and depression. First, is the proposal that the REM state that occurs nightly involves the same spontaneous brainstem process for resolving trauma seen here in the occurrence and disappearance of flashbacks. Second, depression is a form of PTSD, i.e., chronic depression is the autonomic expression of unfinished subcortical attempts to complete the erasure of trauma memory that remains as a residue of interrupted trauma resolution by the same proposed mechanism in REM sleep. The argument for this second hypothesis is derived from the success in alleviating depression by methods that increase “REM pressure” to produce REM rebound.
To begin, the hypothetical description of the REM state is as follows: The REM state begins with the 5 –HT1a mediated suppression of serotonergic neurons in response to the nightly secretion of the endogenous hallucinogen (EH), in concert with supporting noradrenalin action at the locus ceruleus and cholinergic and glutaminergic actions at various other brainstem sites (see table). As in M’s recall and flashbacks, subcortical (and hippocampal) memory substrates are opened via the RaRN mechanism to release stored noxious impulses (trauma) into the cerebral cortical areas for conscious interpretation as physical sensation and dream imagery. Anatomical details of the RaRN mechanism are shown in Figures 9a,b and c for a particular case of pain recall. Associated hippocampal memories are released by the same RaRN model involving suppression of 5-HT neurons of the dorsal and medial raphe nuclei. The resulting cerebral imagery creates a special kind of dream initiated by the release of the noxious components of the hidden traumatic memory. Accordingly, the hypothesis demands the proposal that dreams are not all alike: There are trauma dreams (TDs) that arise from brainstem activity in the REM state and non-trauma dreams (NTDs) that could originate from both REM and non-REM states. NTDs can arise during the non-REM state, or perhaps even the REM state, but TDs arise only from the REM state.
Here, as discussed throughout this monograph, trauma is defined as stored noxious impulses generated by a broad spectrum of viscerally associated incidents that occur during the opening of a sub-cortical memory substrate in response to the secretion of endogenous hallucinogens. Other similar or more intense insults not stored by this process would not qualify as trauma (see 1.5.3 and 1.5.3.1 in the section on Twilight sleep). Stored noxious impulses can take the form of any incident associated with pain of one kind or another. Examples could be in the form of stomach twinges or bradycardia during an incident associated with an urgent mental conflict or global physiological activation from a war injury.
In approaching the question of whether REM sleep might be involved in spontaneous healing, some of its voluminous and confusing history will be confronted, much of which is contrary to this notion. Accordingly, three points will be referenced: 1) REM, though not the only sleep state for dreaming, is postulated to be associated with the kind of dreams that are related to trauma, 2) Earmarks of a REM traumatic dream are the diffuse presence of noxious body sensations arising from sub-cortical storage and the recognition of a specific symbol from the past of the dreamer and 3) REM dreams of the traumatic kind originate from the same brainstem condition of raphe inhibition by the bi-directional RaRN model for opening and closing memory substrates.
More on this second point will be discussed at some length under the section on Twilight Sleep birth, and in the description of Jung’s dream progressions in the page, “Afterwords”. Briefly, the popular clinical therapeutic description of PTSD as “psychological trauma” with its semantic implication of cerebral exclusivity tends to draw attention away from the importance of visceral and somatic pain stored sub-cortically. Yet, brainstem-generated memory may be the primary association for this therapeutic and empirical approach. The pristine physicality of M’s LSD memory recall is from a sub-cortical domain and likely represents a key stimulation for dreams of this kind. Perhaps a better name for “psychological trauma” would be “CNS trauma”.
REM sleep or not, it is generally agreed that dreams do not include recapitulation of physical discomforts or intense emotional states that one would expect from released trauma; they are not an actual re-living of traumatic insults as in M’s case. Apparently, the associations expressed in dreams are exclusive to or minimize visceral somatic sensations and form imagery mainly from cerebral cortical areas. Yet, their bizarre imagery serving as theatre set to distorted plots are much more than the mere recall of the incident as declarative or cognitive memory. Within this fantasyland certain objects and symbols appear that were players in the real incident leading to the memory and are recognizable by the dreamer soon after awakening (Jung, 1943). At this time, some visceral sensations will be felt as details of the real incident are remembered. Indeed, it is often realized after awakening that a rather unpleasant dream occurred during a physical discomfort during the night. These global characteristics identify the dream as one initiated by the release of noxious memory from brainstem memory substrates. These dreams are manifestations of accumulated hidden memories, whose consolidation and release require neural activation within the brainstem.
R4 THE OPPOSITION: MEDICINE AND PSYCHOLOGY
The canons about REM sleep published in 2000 by Solms and those of Vertes and Eastman are found together as fully accessible texts in (http://bbsontime.org/Preprints/OldArchives/bbs.htm). These works provide valuable insights into the origins and functions of the REM state and dreaming, while rejecting the brainstem (the source of REM cycles) as a source of dreams (Solms) that have nothing to do with memory (Vertes and Eastman). These claims are placed into serious question rather neatly by parsing each from the perspective of REM state definitions presented here in the preceding text.
Solms
Solms’ argument that the basal forebrain is not a passive actor as previously thought, but an active area for the generation of dreams may be an important piece in the puzzle for understanding of dreams, together with his finding of the parietal-occipital-temporal nexus as the area for dream interpretation. The problem is the position that dreams, all being of the same kind, can’t originate in one area if they originate in another. Both Jouvet and Solms are correct, but it depends on the kind of dream they are referring to. As hypothesized here, not all dreams are alike. The facts of M’s recall and flashbacks establish the existence and modality of cerebral interpretation of impulses originating in the brainstem, even without the need to invoke the RaRN mechanism. Of course, the images within some sort of distorted plot of an original memory would all require exercise of the cortical areas in the cerebrum, but some dreams would carry associations that must originate from brainstem processes that specialize in the opening of memory substrates. These would be traumatic dreams (TDs), as opposed to non-traumatic dreams (NTDs), whose existence is not in question. NTDs could be unrelated to brainstem or hippocampal storage opening and, as Solms postulates, may well originate in the basal forebrain and even occur during the REM state. However, only the REM state established by brainstem processes shown here can give rise to TDs, whose imagery, again, is manifested in the cerebral cortical areas. Without the suggestion presented here that TDs exist, the conflation of all dreams under the same rubric becomes the central mistake in rejecting the brainstem or basal forebrain as a source in all cases. The possibility of dreams, both originating and not originating in the brainstem, continues as a problem for the following reports.
Vertes and Marshall.
Verdes and Marshall (V&M) may be quite correct in cutting the connection between REM (or NREM) dreaming and memory consolidation, but the evaluation of memory consolidation used in the cited references is incomplete. The survival of learning ability and memory in humans belongs to the category of declarative memory, which is willfully accessed and (apparently) independent of the brainstem (Kandel et al, 2001). Declarative or cognitive memory retrieval exercises cerebral functions and is accessed voluntarily by recalling associations stored in cortical areas under the control of the hippocampus. The maintenance of this kind of learning ability after absorbing different sorts of brainstem injury or living in the absence of REM sleep, even in someone without brainstem function, is not too surprising, since this is only a question of declarative memory that may operate independently of brainstem processes. The brainstem consolidates a different kind of memory defined in this monograph as trauma. Unlike declarative memory, trauma recall is not accessible voluntarily or by any means used in those experiments cited by V&M. What is needed for access to this hidden memory is the subcortical opening of a memory substrate by the interaction between a brainstem 5-HT1a receptor having a high affinity for an endogenous, highly specific agonist. This process would be immune to tricyclic antidepressives or to SSRIs that only encourage moderate increases in 5-HT. Also, the use of SSRIs, as opposed to adding 5-HT, is done in recognition that the sites at issue are highly localized and less accessible to systemic factors. For this same reason, their citation of experiments showing that monoamine oxidase inhibitors (MAOIs) completely abolish REM sleep does not jibe with the kind of parameter measured, i.e., evidence that the subject is unaffected in (declarative) memory or learning.
V&W’s reference to MAOI studies (Vogel et al, 1990; Wyatt et al, 1969) may be the most serious evidence questioning the REM-flashback mechanism proposed here for the nightly onset of the REM state and, indeed, the raphe mechanism itself. As previously discussed and referring to the table above, inhibition of MAO would produce an increase in the monoamines, catecholamine and the tertiary amine, acetylcholine that would not inhibit, but drive REM sleep at each of the REM-ON and REM-OFF sites. This paradox, created by the hypotheses presented in this monograph, would apply to the endogenous hallucinogens, DMT and it’s amine congeners as well, to induce a chronic REM or hallucinogenic state for months, since tolerance is not seen with these drugs (Section 1.9, “What is the endogenous hallucinogen”). Resolution of this koan by invoking the necessity for sequestering these REM-ON and REM-OFF sites away from access to a blood agent is a weaker argument, since other blood components, e.g., LSD do have access. The way out of this is to adopt the present view that the action of MAOI is very unbalanced with respect to the resulting ratios of psychoactive amines. REM sleep is reported to be attributed to noradrenergic and serotonergic blockade to produce this imbalance (Sharpley & Cowen, 1995). In either case, the implication made in the V&M citations, that MAOI action involves the brainstem sites, is questionable.
In summary, cited works within the claims of Vertes and Marshall showing that insults to the brainstem or loss of REM sleep don’t affect memory consolidation is dealt with by two arguments: 1) The criteria used for assessing the effects of REM or brainstem perturbations is not appropriate alone and 2) The results of this monograph reveal the consolidation hidden memory that can reach the cerebral cortex to manifest conscious sensation that might initiate a certain kind of (trauma) dream. There can be no question that a process is inaugurated spontaneously to erase the memory. In addition, evidence for healing a hidden memory is seen in the successful use of a 5-HT1a agonist in PTSD therapy (MAPS.com). This and the similarities between the REM and hallucinogenic states, together with the cerebral awareness of the flashbacks would have to be a consideration in the rejection of memory consolidation by Vertes and Marshall as a component of all dreams.
The value of contributions by Solms, Vertes and Marshall is not diminished by the suggestion made here of more than one source giving rise to dreams. These authors may be correct if their reference is restricted to non-trauma related dreams in their respective positions about basal forebrain sources and memory consolidation. These authors have additional critics on similar issues (Bednar, 2003). Therefore, the extension of these experimental results of spontaneous flashbacks as signs of trauma erasure to the mechanism of REM sleep remains as an hypothesis worthy of testing. The appearance of the agonist in the blood as the REM state begins and ends could be managed (see "Testing the Hypothesis). Just as brainstem events documented here lead to conscious awareness of released memory impulses, the release of stored impulses by the same brainstem events in the REM state reach their conscious manifestation as dreams from cerebral processes. The confirmation or falsification of these predictions is possible by testing based on neurobiological methods with the 5-HT1a receptor as target and the use of functional imaging (fMRI) and PET, as outlined in the same section on testing. The anticipation is that REM dreams are evidence of a healing function repeated nightly to rid the dreamer of noxious impulses stored according to the RaRN model. Whether or not the 5-HT1a receptor really belongs to the raphe nuclei or to another thus far unknown source is a valid question, but irrelevant to the fact that cerebral transformation into conscious sensation or dream imagery can originate from 1a-ligand binding within the brainstem.
R5 A PROPOSED ETIOLOGY OF DEPRESSION
A major finding of this monograph is the evidence of a memory substrate controlled by a 5-HT1a agonist for consolidation or release of traumatic impulses. The model generated from this evidence has been exemplified by known anatomical and neural interactions and accounts for a number of general observations in the areas of hallucinogen research and, notably, in the area of posttraumatic stress disorder (PTSD). The RaRN model accounts for the frequent persistence of noxious flashbacks and PTSD in the clinical setting.
Thus, a natural outcome of combining the REM-flashback hypothesis with clinical studies on REM sleep is the proposal that chronic depression is an external manifestation of autonomic attempts to complete the resolution of a trauma memory whose resolution was not completed during REM sleep. In other words, the daytime condition of certain depressive individuals is a chronic form of posttraumatic stress disorder (PTSD). The basis for this hypothesis is the documented success of treating depression by methods that increase REM pressure to induce “REM rebound” (Bednar, 2003). In these cases, the patient is awakened each time the REM state is detected, either by EEG theta waves or by rapid eye movement. This awakening is repeated for one or two nights. If the patient is undisturbed the third night, his REM latency will be normal and dreams may be intense as the patient “catches up” on REM time. The next day the depression is gone. Although this approach does not produce lasting relief from depression, the observation is fairly reliable and was suggested by earlier observations that loss of sleep often produces an improvement in depression, while napping does the opposite. The rationale for using these clinical observations is as follows: In some depressives the brainstem mechanism of opening memory substrates containing noxious impulses is insufficient, owing to insufficient REM pressure to reach a threshold for agonist secretion or for the activation of second messengers as described in later sections to hyperpolarize functional target areas. Increasing REM pressure is required to reach this threshold and release the hidden memory into cortical dream interpretation. The question as to what neural processes give rise to REM pressure is an intriguing one. The autonomic system is keeping track of jobs unfinished. As mentioned in "Afterword," a "reference" is continually monitored and, as in M's flashbacks, personal issues in dreams diminish with time to some end-point (Jung, 1943).
These conjectures give rise to two considerations, one about the role of REM dreaming and the other, the relation between REM sleep and depression. First, is the proposal that the REM state that occurs nightly involves the same spontaneous brainstem process for resolving trauma seen here in the occurrence and disappearance of flashbacks. Second, depression is a form of PTSD, i.e., chronic depression is the autonomic expression of unfinished subcortical attempts to complete the erasure of trauma memory that remains as a residue of interrupted trauma resolution by the same proposed mechanism in REM sleep. The argument for this second hypothesis is derived from the success in alleviating depression by methods that increase “REM pressure” to produce REM rebound.
To begin, the hypothetical description of the REM state is as follows: The REM state begins with the 5 –HT1a mediated suppression of serotonergic neurons in response to the nightly secretion of the endogenous hallucinogen (EH), in concert with supporting noradrenalin action at the locus ceruleus and cholinergic and glutaminergic actions at various other brainstem sites (see table). As in M’s recall and flashbacks, subcortical (and hippocampal) memory substrates are opened via the RaRN mechanism to release stored noxious impulses (trauma) into the cerebral cortical areas for conscious interpretation as physical sensation and dream imagery. Anatomical details of the RaRN mechanism are shown in Figures 9a,b and c for a particular case of pain recall. Associated hippocampal memories are released by the same RaRN model involving suppression of 5-HT neurons of the dorsal and medial raphe nuclei. The resulting cerebral imagery creates a special kind of dream initiated by the release of the noxious components of the hidden traumatic memory. Accordingly, the hypothesis demands the proposal that dreams are not all alike: There are trauma dreams (TDs) that arise from brainstem activity in the REM state and non-trauma dreams (NTDs) that could originate from both REM and non-REM states. NTDs can arise during the non-REM state, or perhaps even the REM state, but TDs arise only from the REM state.
Here, as discussed throughout this monograph, trauma is defined as stored noxious impulses generated by a broad spectrum of viscerally associated incidents that occur during the opening of a sub-cortical memory substrate in response to the secretion of endogenous hallucinogens. Other similar or more intense insults not stored by this process would not qualify as trauma (see 1.5.3 and 1.5.3.1 in the section on Twilight sleep). Stored noxious impulses can take the form of any incident associated with pain of one kind or another. Examples could be in the form of stomach twinges or bradycardia during an incident associated with an urgent mental conflict or global physiological activation from a war injury.
In approaching the question of whether REM sleep might be involved in spontaneous healing, some of its voluminous and confusing history will be confronted, much of which is contrary to this notion. Accordingly, three points will be referenced: 1) REM, though not the only sleep state for dreaming, is postulated to be associated with the kind of dreams that are related to trauma, 2) Earmarks of a REM traumatic dream are the diffuse presence of noxious body sensations arising from sub-cortical storage and the recognition of a specific symbol from the past of the dreamer and 3) REM dreams of the traumatic kind originate from the same brainstem condition of raphe inhibition by the bi-directional RaRN model for opening and closing memory substrates.
More on this second point will be discussed at some length under the section on Twilight Sleep birth, and in the description of Jung’s dream progressions in the page, “Afterwords”. Briefly, the popular clinical therapeutic description of PTSD as “psychological trauma” with its semantic implication of cerebral exclusivity tends to draw attention away from the importance of visceral and somatic pain stored sub-cortically. Yet, brainstem-generated memory may be the primary association for this therapeutic and empirical approach. The pristine physicality of M’s LSD memory recall is from a sub-cortical domain and likely represents a key stimulation for dreams of this kind. Perhaps a better name for “psychological trauma” would be “CNS trauma”.
REM sleep or not, it is generally agreed that dreams do not include recapitulation of physical discomforts or intense emotional states that one would expect from released trauma; they are not an actual re-living of traumatic insults as in M’s case. Apparently, the associations expressed in dreams are exclusive to or minimize visceral somatic sensations and form imagery mainly from cerebral cortical areas. Yet, their bizarre imagery serving as theatre set to distorted plots are much more than the mere recall of the incident as declarative or cognitive memory. Within this fantasyland certain objects and symbols appear that were players in the real incident leading to the memory and are recognizable by the dreamer soon after awakening (Jung, 1943). At this time, some visceral sensations will be felt as details of the real incident are remembered. Indeed, it is often realized after awakening that a rather unpleasant dream occurred during a physical discomfort during the night. These global characteristics identify the dream as one initiated by the release of noxious memory from brainstem memory substrates. These dreams are manifestations of accumulated hidden memories, whose consolidation and release require neural activation within the brainstem.
R4 THE OPPOSITION: MEDICINE AND PSYCHOLOGY
The canons about REM sleep published in 2000 by Solms and those of Vertes and Eastman are found together as fully accessible texts in (http://bbsontime.org/Preprints/OldArchives/bbs.htm). These works provide valuable insights into the origins and functions of the REM state and dreaming, while rejecting the brainstem (the source of REM cycles) as a source of dreams (Solms) that have nothing to do with memory (Vertes and Eastman). These claims are placed into serious question rather neatly by parsing each from the perspective of REM state definitions presented here in the preceding text.
Solms
Solms’ argument that the basal forebrain is not a passive actor as previously thought, but an active area for the generation of dreams may be an important piece in the puzzle for understanding of dreams, together with his finding of the parietal-occipital-temporal nexus as the area for dream interpretation. The problem is the position that dreams, all being of the same kind, can’t originate in one area if they originate in another. Both Jouvet and Solms are correct, but it depends on the kind of dream they are referring to. As hypothesized here, not all dreams are alike. The facts of M’s recall and flashbacks establish the existence and modality of cerebral interpretation of impulses originating in the brainstem, even without the need to invoke the RaRN mechanism. Of course, the images within some sort of distorted plot of an original memory would all require exercise of the cortical areas in the cerebrum, but some dreams would carry associations that must originate from brainstem processes that specialize in the opening of memory substrates. These would be traumatic dreams (TDs), as opposed to non-traumatic dreams (NTDs), whose existence is not in question. NTDs could be unrelated to brainstem or hippocampal storage opening and, as Solms postulates, may well originate in the basal forebrain and even occur during the REM state. However, only the REM state established by brainstem processes shown here can give rise to TDs, whose imagery, again, is manifested in the cerebral cortical areas. Without the suggestion presented here that TDs exist, the conflation of all dreams under the same rubric becomes the central mistake in rejecting the brainstem or basal forebrain as a source in all cases. The possibility of dreams, both originating and not originating in the brainstem, continues as a problem for the following reports.
Vertes and Marshall.
Verdes and Marshall (V&M) may be quite correct in cutting the connection between REM (or NREM) dreaming and memory consolidation, but the evaluation of memory consolidation used in the cited references is incomplete. The survival of learning ability and memory in humans belongs to the category of declarative memory, which is willfully accessed and (apparently) independent of the brainstem (Kandel et al, 2001). Declarative or cognitive memory retrieval exercises cerebral functions and is accessed voluntarily by recalling associations stored in cortical areas under the control of the hippocampus. The maintenance of this kind of learning ability after absorbing different sorts of brainstem injury or living in the absence of REM sleep, even in someone without brainstem function, is not too surprising, since this is only a question of declarative memory that may operate independently of brainstem processes. The brainstem consolidates a different kind of memory defined in this monograph as trauma. Unlike declarative memory, trauma recall is not accessible voluntarily or by any means used in those experiments cited by V&M. What is needed for access to this hidden memory is the subcortical opening of a memory substrate by the interaction between a brainstem 5-HT1a receptor having a high affinity for an endogenous, highly specific agonist. This process would be immune to tricyclic antidepressives or to SSRIs that only encourage moderate increases in 5-HT. Also, the use of SSRIs, as opposed to adding 5-HT, is done in recognition that the sites at issue are highly localized and less accessible to systemic factors. For this same reason, their citation of experiments showing that monoamine oxidase inhibitors (MAOIs) completely abolish REM sleep does not jibe with the kind of parameter measured, i.e., evidence that the subject is unaffected in (declarative) memory or learning.
V&W’s reference to MAOI studies (Vogel et al, 1990; Wyatt et al, 1969) may be the most serious evidence questioning the REM-flashback mechanism proposed here for the nightly onset of the REM state and, indeed, the raphe mechanism itself. As previously discussed and referring to the table above, inhibition of MAO would produce an increase in the monoamines, catecholamine and the tertiary amine, acetylcholine that would not inhibit, but drive REM sleep at each of the REM-ON and REM-OFF sites. This paradox, created by the hypotheses presented in this monograph, would apply to the endogenous hallucinogens, DMT and it’s amine congeners as well, to induce a chronic REM or hallucinogenic state for months, since tolerance is not seen with these drugs (Section 1.9, “What is the endogenous hallucinogen”). Resolution of this koan by invoking the necessity for sequestering these REM-ON and REM-OFF sites away from access to a blood agent is a weaker argument, since other blood components, e.g., LSD do have access. The way out of this is to adopt the present view that the action of MAOI is very unbalanced with respect to the resulting ratios of psychoactive amines. REM sleep is reported to be attributed to noradrenergic and serotonergic blockade to produce this imbalance (Sharpley & Cowen, 1995). In either case, the implication made in the V&M citations, that MAOI action involves the brainstem sites, is questionable.
In summary, cited works within the claims of Vertes and Marshall showing that insults to the brainstem or loss of REM sleep don’t affect memory consolidation is dealt with by two arguments: 1) The criteria used for assessing the effects of REM or brainstem perturbations is not appropriate alone and 2) The results of this monograph reveal the consolidation hidden memory that can reach the cerebral cortex to manifest conscious sensation that might initiate a certain kind of (trauma) dream. There can be no question that a process is inaugurated spontaneously to erase the memory. In addition, evidence for healing a hidden memory is seen in the successful use of a 5-HT1a agonist in PTSD therapy (MAPS.com). This and the similarities between the REM and hallucinogenic states, together with the cerebral awareness of the flashbacks would have to be a consideration in the rejection of memory consolidation by Vertes and Marshall as a component of all dreams.
The value of contributions by Solms, Vertes and Marshall is not diminished by the suggestion made here of more than one source giving rise to dreams. These authors may be correct if their reference is restricted to non-trauma related dreams in their respective positions about basal forebrain sources and memory consolidation. These authors have additional critics on similar issues (Bednar, 2003). Therefore, the extension of these experimental results of spontaneous flashbacks as signs of trauma erasure to the mechanism of REM sleep remains as an hypothesis worthy of testing. The appearance of the agonist in the blood as the REM state begins and ends could be managed (see "Testing the Hypothesis). Just as brainstem events documented here lead to conscious awareness of released memory impulses, the release of stored impulses by the same brainstem events in the REM state reach their conscious manifestation as dreams from cerebral processes. The confirmation or falsification of these predictions is possible by testing based on neurobiological methods with the 5-HT1a receptor as target and the use of functional imaging (fMRI) and PET, as outlined in the same section on testing. The anticipation is that REM dreams are evidence of a healing function repeated nightly to rid the dreamer of noxious impulses stored according to the RaRN model. Whether or not the 5-HT1a receptor really belongs to the raphe nuclei or to another thus far unknown source is a valid question, but irrelevant to the fact that cerebral transformation into conscious sensation or dream imagery can originate from 1a-ligand binding within the brainstem.
R5 A PROPOSED ETIOLOGY OF DEPRESSION
A major finding of this monograph is the evidence of a memory substrate controlled by a 5-HT1a agonist for consolidation or release of traumatic impulses. The model generated from this evidence has been exemplified by known anatomical and neural interactions and accounts for a number of general observations in the areas of hallucinogen research and, notably, in the area of posttraumatic stress disorder (PTSD). The RaRN model accounts for the frequent persistence of noxious flashbacks and PTSD in the clinical setting.
Thus, a natural outcome of combining the REM-flashback hypothesis with clinical studies on REM sleep is the proposal that chronic depression is an external manifestation of autonomic attempts to complete the resolution of a trauma memory whose resolution was not completed during REM sleep. In other words, the daytime condition of certain depressive individuals is a chronic form of posttraumatic stress disorder (PTSD). The basis for this hypothesis is the documented success of treating depression by methods that increase REM pressure to induce “REM rebound” (Bednar, 2003). In these cases, the patient is awakened each time the REM state is detected, either by EEG theta waves or by rapid eye movement. This awakening is repeated for one or two nights. If the patient is undisturbed the third night, his REM latency will be normal and dreams may be intense as the patient “catches up” on REM time. The next day the depression is gone. Although this approach does not produce lasting relief from depression, the observation is fairly reliable and was suggested by earlier observations that loss of sleep often produces an improvement in depression, while napping does the opposite. The rationale for using these clinical observations is as follows: In some depressives the brainstem mechanism of opening memory substrates containing noxious impulses is insufficient, owing to insufficient REM pressure to reach a threshold for agonist secretion or for the activation of second messengers as described in later sections to hyperpolarize functional target areas. Increasing REM pressure is required to reach this threshold and release the hidden memory into cortical dream interpretation. The question as to what neural processes give rise to REM pressure is an intriguing one. The autonomic system is keeping track of jobs unfinished. As mentioned in "Afterword," a "reference" is continually monitored and, as in M's flashbacks, personal issues in dreams diminish with time to some end-point (Jung, 1943).
1.4.5 SOME ADDITIONAL REFERENCES FOR REM SLEEP
Gillin JC, Sohn J-W, Stahl SM, Lardon M, Kelsoe J, Rapaport M, Ruiz C and Golsham S. (1996) Ipsapirone, a 5-HT1a agonist, suppresses REM sleep equally in unmidicated depressived patients and normal controls. Neuropharmacology 15, 109-15.
Wilson SJ, Bailey JE, Rich AS, Nash J, Adrover M, Tournoux A and Nutt DJ. (2005) The use of sleep measures to compare a neew 5HT1a agonist with buspirone in humans. J. Psychopharmacoogy 12, 609-13.
Monti JM and Monti D. (2000) Role of dorsal raphe nucleus serotonin 5-HT1a receptor in the regulation of REM sleep. Life Sciences 66, 1999-2012
See other citations in the REFERENCE page of this website.
Gillin JC, Sohn J-W, Stahl SM, Lardon M, Kelsoe J, Rapaport M, Ruiz C and Golsham S. (1996) Ipsapirone, a 5-HT1a agonist, suppresses REM sleep equally in unmidicated depressived patients and normal controls. Neuropharmacology 15, 109-15.
Wilson SJ, Bailey JE, Rich AS, Nash J, Adrover M, Tournoux A and Nutt DJ. (2005) The use of sleep measures to compare a neew 5HT1a agonist with buspirone in humans. J. Psychopharmacoogy 12, 609-13.
Monti JM and Monti D. (2000) Role of dorsal raphe nucleus serotonin 5-HT1a receptor in the regulation of REM sleep. Life Sciences 66, 1999-2012
See other citations in the REFERENCE page of this website.
