- Email: [email protected]
The conversational brain: Fronto-hippocampal interaction and disconnection
Medical Hypotheses (2006) 67, 759–764
http://intl.elsevierhealth.com/journals/mehy
The conversational brain: Fronto-hippocampal interaction and disconnection J. David Johnson 2021 Jolly Road, Okemos, MI 48864, United States Received 6 April 2006; accepted 10 April 2006
Summary The paper promotes the view that the alert brain alternates between operating in an action mode, based on frontal lobe function, and a receptive mode, involving cholinergic system activity. Their alternation forms a conversation with the environment. It is hypothesized that competition between the modes centers on control over excitability of neurons in the CA1 field of the hippocampus. Increased excitability enhances the flow of hippocampal output through the subiculum resulting in support for frontal lobe function and the action mode. Decreased excitability, on the other hand, reduces this output and that support, leading to a disconnection between frontal lobes and hippocampus. At the same time, correlated cholinergic activity enhances receptive mode processes, indicated by the occurrence of the hippocampal theta rhythm. It is suggested that the hypothesis provides a conceptual framework for considering various phenomena including REM sleep, schizophrenia, and hypnosis. In REM sleep the receptive mode remains dominant as cholinergic activity supports the hippocampal integration of experience into a composite view of reality. In schizophrenia, the action and receptive modes are not properly coordinated because of a dysfunction in anterior hippocampal output. And hypnosis might be seen as a process in which conditions and suggestions are able to induce in some people a prolonged occurrence of the receptive mode allowing a normal view of reality to be altered. c 2006 Elsevier Ltd. All rights reserved.
Introduction The frontal lobes are generally seen as providing the basis for the brain’s executive ability [1]. Their function underlies the initiation, control and maintenance of planned, goal-oriented behavior. Furthermore, they are responsible for mechanisms that protect this behavior, possibly by suppressing activity that would interfere with its realization [2]. However, it seems that there are conditions where the expression of executive control is unE-mail address: [email protected].
needed and even disruptive. These fall into two opposite categories. On the one hand, when behavior is highly practiced and the environment is predictable, it may be best to allow the practiced behavior to prevail over ‘‘second guessing’’ by executive management. On the other hand, when the environment suddenly becomes uncertain or novel experiences appear, it may be adaptive to suspend ongoing planned behavior because such behavior may no longer be appropriate under the new conditions. The need to adjust to changing conditions suggests that there exists a level of control above the executive power of the frontal lobes that is
0306-9877/$ - see front matter c 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2006.04.031
760 able to undermine or interrupt behavior under their direction. It would enable the brain to switch from an executive or action mode to a receptive mode, a mode favoring investigation. As the individual encountered novel experiences and developed ways to deal with them, these modes would alternate back and forth. In effect, interaction with the environment would be conversational, alternating between receptive (e.g., listening) and action (e.g., speaking). This paper hypothesizes that the ability to control the alternation between action and receptive modes rests with the hippocampus. Hippocampal output via the subiculum is seen as a facilitator of executive functions of the frontal lobes thus supporting the action mode. On the other hand, restriction of that output results in an attenuation of those functions and occurs together with a facilitation of the receptive mode. During the action mode, the frontal lobes and hippocampus interact and work together in an integrated fashion, but in the receptive mode there is a disconnection between frontal lobes and hippocampus. The hypothesis is presented as a framework for interpreting the relation between executive and receptive brain functions in various situations. These include the normal waking alert state, REM sleep, schizophrenia, and hypnosis.
Hippocampal theta and the receptive mode The hypothesis is based on the idea that there exist certain brain functions that provide executive control and others that encourage the reception and processing of new information. It is assumed that executive control is a product of frontal lobe function. Receptive processes will be associated with the activity of cholinergic systems. There is strong evidence implicating the cholinergic systems with the creation of conditions that favor the reception of information. Studies of the brainstem cholinergic system have shown that its activity facilitates the relay of sensory signals through the thalamus [3]. Projections to the cortex from the basal forebrain cholinergic system appear to enhance the receptivity of cortical neurons [4]. The medial septal nucleus is part of the basal forebrain system and it projects to the hippocampus. The activity of this projection is widely seen as an important cause of the hippocampal theta rhythm [5]. Numerous studies have been directed at understanding the significance of this rhythm. Many of
Johnson these observed the behavior that accompanies the appearance of hippocampal theta. In a review of such studies in several species, Miller [5] concluded that theta was most likely to appear at times when it was important for the organism to be gathering information about its environment. This was illustrated in a study by Kemp and Kaada with cats [6]. They observed that during waking, maximal hippocampal theta activity was associated with ‘‘fixed staring and visual searching behavior’’. These general observational findings are consistent with the idea that hippocampal theta signals the occurrence of a receptive state. More specific evidence for this can be seen in a study by Tesche and Karhu [7] of human hippocampal activity during a working memory task. They observed that the presentation of the to be remembered stimuli invoked theta, and after several trials theta onset in the right hippocampus began to anticipate the appearance of these stimuli. A burst of hippocampal theta also occurred in response to the probe event that provided the subject the information to properly categorize the stimulus set. These results suggest that theta occurs when the individual prepares to receive information and when the information is being processed. We would associate both conditions with the receptive mode.
Mechanisms underlying frontohippocampal interaction So far it has been proposed that there exists an action mode based on frontal lobe function and a receptive mode based on cholinergic system activity and associated particularly with hippocampal theta activity. The hypothesis includes the idea that these modes compete with each other for dominance. There is evidence to suggest that the competition centers on the control over hippocampal output flowing from the CA1 field and subiculum. It will be argued that restricting this output favors the receptive mode whereas its facilitation benefits the action mode. The evidence begins with experiments showing that during the occurrence of hippocampal theta, output projecting through the subiculum from CA1 is reduced. Winson [8] studied the flow of activity through the hippocampus during several states including theta during waking. He found that at CA1, the last stage in this flow, activity was consistently diminished during theta either in waking or in REM sleep. Other researchers found similar suppression of CA1 spikes in experiments that specifically invoked theta in several ways [9,10]. The
The conversational brain: Fronto-hippocampal interaction and disconnection suppression of CA1 excitability during theta has been shown to originate in reticular formation activation reaching the hippocampus via supramammillary and septal regions [11]. But there is also evidence of an opposite influence on CA1 excitability. It has been shown that input to the hippocampus from the thalamic nucleus reuniens can increase CA1 excitability [12]. McKenna and Vertes [13] observed that the prefrontal cortex provides a major input to the nucleus reuniens and suggested that the nucleus supports an important route of communication between prefrontal cortex and the hippocampus. Thus frontal lobe activity could result in a facilitation of the flow of hippocampal output through CA1 and the subiculum. The significance of a competition for control over hippocampal output can be seen in studies of the effects of that output on target structures. Its effect on the nucleus accumbens is particularly interesting. Grace and colleagues have studied this effect in the context of a circuit that goes from accumbens to ventral pallidum to mediodorsal thalamic nucleus to prefrontal cortex and back to accumbens [14]. They observed that hippocampal output to the accumbens enabled flow through this circuit by allowing activation of accumbens cells by prefrontal cortex to be relayed on to the ventral pallidum. In discussing these results they suggested that the hippocampus is therefore able to influence the level of activity in the prefrontal cortex. By this means, frontal lobe function could be attenuated when CA1 excitability is limited during theta, but could be enhanced when activity originating in the prefrontal cortex via nucleus reuniens leads to increased CA1 excitability and hippocampal output. The circuit implicated in the model of O’Donnell and Grace [14] could be characterized as providing motivational support for frontal lobe function [15]. The parahippocampal gyrus provides another basis for interaction between hippocampus and prefrontal cortex that may influence the specific nature of frontal lobe activity. The gyrus receives output from the subiculum and relays it on to many cortical areas including prefrontal cortex [16]. Previously it was suggested that the wide spread projection mediated by the parahippocampal gyrus supports the imposition of a contextual influence upon target structures [17]. If that is true, control of hippocampal output at CA1 might not only determine the motivation for pursuing goal-oriented behavior but might also have an effect on processes shaping the direction of this behavior. In short, it appears that during hippocampal theta and the occurrence of the receptive mode, hip-
761
pocampal output is restricted at CA1. This creates a disconnection between hippocampus and frontal lobes that may lead to an attenuation of frontal lobe function and suspension of the action mode. On the other hand, when frontal lobe activity increases enabling the action mode, the frontal lobes and hippocampus are able to interact. The presence of the action mode may be reinforced as increased CA1 excitability allows hippocampal output to support frontal lobe function.
Interaction and disconnection during normal waking Previous papers presented a view of hippocampal function which centered on the idea that the hippocampus alternates between a theta mode and a non-theta mode [17–19]. During the theta mode, input to the hippocampus is unrestricted but output through the subiculum is restricted. While in this mode, a variety of information from throughout the brain becomes integrated in the hippocampus. This results in the formation of context memory which is a composite picture of the reliable aspects of current experience. This memory serves as an integrated set of expectations about the environment, a worldview. By comparison, during the non-theta mode, input to the hippocampus is restricted while output through the subiculum is unrestricted. The pattern of this output is based on the context memory formed during the theta mode and possibly becomes copied into the parahippocampal gyrus. The output is seen as having two major functions. First, it creates throughout the neocortex a scaffold on which the patterns of specific episodes may become attached. These associations provide the basis for retrieval of the memory of episodes. Secondly, the output creates a widespread contextual influence. The function of that influence is to cause perception, emotion and behavior to be consistent with the underlying context memory. The influence was described as providing a worldview and reality orientation. Orientation in space was linked to output from the dorsal hippocampus whereas emotional orientation was attributed to output from the anterior hippocampus. The concepts of receptive and action modes described here correspond to and are extensions of the theta and non-theta modes described previously. When the receptive mode is present during normal waking, it represents a naturally occurring disconnection between hippocampal and frontal lobe functions. It is expected that this sort of
762 disconnection would be most prominent when individuals are young or later when they are placed in a novel environment, i.e., when there is a maximum amount of new information that should be acquired in order to adapt. But as individuals mature and conditions become familiar, much less time would be spent in the receptive mode and the action mode would become increasingly dominant during alert waking, especially in species with highly developed frontal lobes. This may explain why it has been difficult to observe hippocampal theta in adult primates [20], and when it is seen, it occurs in brief bursts [7]. In addition to degree of experience and species differences, other factors could determine the extent to which the action or receptive mode is dominant. For instance, if the novel event were immediately perceived as threatening, individuals may forgo extensive investigation and quickly begin deriving plans for action. But if the novel event were merely a curiosity and promoted relaxation, they might remain longer in the receptive mode. Furthermore, individuals might naturally vary in the extent to which one or the other mode prevailed. Some might maintain executive control almost continuously whereas others might more readily suspend this control.
Other cases of interaction and disconnection Previously the phenomena of REM sleep and of schizophrenia were discussed in terms of hippocampal function [18,19]. Here those discussions will be extended to consider the phenomena in terms of the concepts of action and receptive modes. Then the discussion will turn to a consideration of the nature of hypnosis from the perspective of this conceptual framework.
REM sleep Periods of REM sleep coincide with the appearance of hippocampal theta [21]. Therefore we would expect that REM sleep is a time in which the receptive mode remains dominant and frontal lobes are disconnected from the hippocampus. Evidence of the disconnection can be found in measures of brain activity during REM. Braun et al. [22] observed a marked reduction in frontal lobe activity during REM when compared either to waking or slow wave sleep. At the same time, the hippocampus and parahippocampal cortices showed increased activation.
Johnson REM sleep is also commonly associated with the occurrence of vivid dreaming [23]. A reflection on the nature of dreams suggests support for the idea that they occur when the action mode has been replaced by the receptive mode. Dreams lack the planning and executive control that is characteristic of the action mode. Instead, they often seem to be series of experiences that we passively observe or respond to reflexively, as would be expected when one is in the receptive mode. According to the present view, REM sleep is essentially a period of input processing where the input is a variety of information flowing into the hippocampus. It has been hypothesized that during REM sleep conditions are ideal for developing a composite memory of current experience [18]. The reduction in frontal lobe activity during this time was said to allow a free flow of information, free of the bias of executive control.
Schizophrenia Over the years various hypotheses have been offered to explain the cause or causes of schizophrenia. Recent measures of brain activity in schizophrenics suggest that dysfunction within the medial temporal lobe, the anterior hippocampus in particular, may be an underlying cause of this disease [24]. This idea has gained support from studies showing that hippocampal lesions in neonatal rats can cause schizophrenia-like symptoms when they mature [25]. Dysfunction of the anterior hippocampus would create an unnatural disconnection between the hippocampus and the targets of its output. Previously, it was argued that fundamental schizophrenic symptoms, such as inappropriate affect and splitting of associations, were caused by a failure of anterior hippocampal output to control its target structures [19]. Inappropriate affect was the result of a lack of a contextual influence upon emotional processes. The splitting of associations was caused by a failure of this output to support frontal lobe control over the flow of associations. When viewed in terms of the idea of competing action and receptive modes, schizophrenia could be seen as a condition in which these modes are not properly coordinated. Dysfunction of the anterior hippocampus could create a lasting disconnection between hippocampal and frontal lobe activity. This may account for the finding of reduced frontal lobe activity in some schizophrenics [26]. O’Donnell and Grace [14] suggested that their hippocampal gating model might explain the cause of this hypofrontality. However, not all schizophrenics show decreased frontal lobe activity. For
The conversational brain: Fronto-hippocampal interaction and disconnection example, Laurie et al. [27] observed brain activity in schizophrenics and normals and reported both absolute levels and correlations between frontal and temporal lobe activity. They found that levels of frontal lobe activity did not differ between normals and patients. But activities of frontal and temporal lobes were significantly less correlated in schizophrenics. These results suggest that frontal lobe activity can achieve a normal level without the support of anterior hippocampal output, but that this output may be necessary for proper coordination of the functions of frontal and temporal lobes including the control of competition between action and receptive processes.
Hypnosis The term hypnosis refers to a condition in which it is said that sensations, perceptions, emotions, and behavior can be altered by suggestion. However, the existence of such a state remains controversial. The present paper does not take a position on the controversy. But it does suggest that the concepts of action and receptive modes provide a useful framework for discussing hypnosis and point to a way to test for its existence. Critics of the existence of hypnosis maintain that the subjects of a hypnotic session are behaving in a goal directed manner and are trying to satisfy the expectations and wishes of the hypnotist [28]. In terms of the current framework, the subjects remain predominantly in the action mode during and after a session. Frontal lobe functions remain in control throughout the experience. However, it is possible to imagine a different scenario. Conceivably, for some people the stimuli and instructions presented during hypnotic induction might elicit the receptive mode such that it becomes dominant and sustained. If that were to occur, we would expect that persons would accept their experience uncritically (as happens during REM sleep dreaming) due to an attenuation of frontal lobe function. And the information they receive in this state, including hypnotic suggestions, could be incorporated in the worldview that is formed when in the receptive/theta mode [18,19]. As a result, persons could emerge from a session with an altered view of reality that is enforced when the action mode regains control. To establish that the events in this scenario could actually occur, one needs evidence from measures of brain function. It is important to distinguish two kinds of evidence. The first would be measures of the effects of hypnotic suggestions. These results could be as varied as the suggestions
763
can be. For instance, hypnotic suggestions intended to reduce conflict were shown to decrease activity in the anterior cingulate cortex [29] whereas suggestions that pain will be experienced lead to increased activity in this cortex [30]. The second kind of evidence would be measures of underlying conditions that allow a person to be hypnotized. The conceptual framework promoted here predicts the results one would expect if hypnosis exists. There should be evidence associating hypnotizability with decreased frontal lobe function (which would weaken the control of the action mode) and increased occurrence of hippocampal theta, especially during hypnotic induction (signaling the dominance of the receptive mode). A reduction in measures of frontal lobe function in highly hypnotizable subjects has already been reported [31]. It remains to be seen whether these subjects also show an above average increase in hippocampal theta during a hypnotic session.
Conclusion This paper promotes the idea that brain function alternates between a receptive mode and an action mode. The receptive mode favors the gathering of information whereas the action mode supports the generation of goal-oriented behavior. The action mode relies on the activity of recently evolved neocortex of the frontal lobes while the receptive mode is based on much older cholinergic systems. From an evolutionary perspective their interaction is a competition between newer and older brain functions. If the difficulty to observe hippocampal theta in primates is any guide, it appears that the action mode in mammals is becoming the stronger competitor in those with larger frontal lobes [20]. However, there will always be a need to gather and process information from our environment. Executive control would be acting blindly if receptive mode processes were entirely overridden by the action mode. So we would expect the receptive mode to remain competitive. If it were shown that hypnotizability really exists and is a trait that allows the receptive mode to be easily invoked, we could speculate that this trait is adaptive in humans in as much as it enables the receptive mode to remain competitive in the face of a highly assertive action mode.
References [1] Goldberg E. The executive brain. Oxford: Oxford University Press; 2001.
764 [2] Fuster JM. The prefrontal cortex: anatomy, physiology and neuropsychology of the frontal lobes. 3rd ed. Philadelphia: Lippincott-Raven; 1997. [3] Steriade M, Buzsaki G. Parallel activation of thalamic and cortical neurons by brainstem and basal forebrain cholinergic systems. In: Steriade M, Biesold D, editors. Brain cholinergic systems. Oxford: Oxford University Press; 1990. p. 1–62. [4] McCormick DA. Cellular mechanisms of cholinergic control of neocortical and thalamic neuronal excitability. In: Steriade M, Biesold D, editors. Brain cholinergic systems. Oxford: Oxford University Press; 1990. p. 236–64. [5] Miller R. Cortico-hippocampal interplay and the representation of contexts. Berlin: Springer-Verlag; 1991. p. 83–85. [6] Kemp IR, Kaada BR. The relation of hippocampal theta activity to arousal, attentive behaviour and somato-motor movements in unrestrained cats. Brain Res 1975;95:323–434. [7] Tesche CD, Karhu J. Theta oscillations index human hippocampal activation during a working memory task. Proc Natl Acad Sci USA 2000;97:919–24. [8] Winson J. Neuronal transmission through the hippocampus: dependence on behavioral state. In: Reinoso-Suarez F, Ajmone-Marsan C, editors. Cortical integration. New York: Raven Press; 1984. p. 131–46. [9] Herreras O, Solis JM, Munoz MD, Martin del Rio R, Lerma J. Sensory synapsis modulation of hippocampal transmission. I. Opposite effects on CA1 and dentate gyrus. Brain Res 1988;461:290–302. [10] Vinogradova OS, Brazhnik ES, Kitchigina VF, Stafekhina VS. Acetylcholine, theta-rhythm and activity of hippocampal neurons in the rabbit – IV. Sensory stimulation. Neuroscience 1993;53:993–1007. [11] Jiang F, Khanna S. Reticular stimulation evokes suppression of CA1 synaptic responses and generation of theta through separate mechanisms. Eur J Neurosci 2004;19:295–308. [12] Bertram EH, Zhang DX. Thalamic excitation of hippocampal CA1 neurons: a comparison with the effects of CA3 stimulation. Neuroscience 1999;92:15–26. [13] McKenna JT, Vertes RP. Afferent projections to nucleus reuniens of the thalamus. J Comp Neurol 2004;480:115–42. [14] O’Donnell P, Grace AA. Synaptic interactions among excitatory afferents to nucleus accumbens neurons: hippocampal gating of prefrontal cortical input. J Neurosci 1995;15:3622–39. [15] Chow TW, Cummings JL. Frontal-subcortical circuits. In: Miller BL, Cummings JL, editors. The human frontal lobes: functions and disorders. New York: The Guilford Press; 1999. p. 3–26.
Johnson [16] VanHoesen GW. The parahippocampal gyrus: new observations regarding its cortical connections in the monkey. Trends Neurosci 1982;5:345–50. [17] Johnson JD. Episodic memory and the hippocampus: another view. Med Hypoth 2004;63:963–7. [18] Johnson JD. REM sleep and the development of context memory. Med Hypoth 2005;64:499–504. [19] Johnson JD. Dysfunction of the anterior hippocampus: the cause of fundamental schizophrenic symptoms? Med Hypoth 2005;65:55–60. [20] Stewart M, Fox SE. Hippocampal theta activity in monkeys. Brain Res 1991;538:59–63. [21] Vertes RP, Kocsis B. Brainstem-diencephalo-septohippocampal systems controlling the theta rhythm of the hippocampus. Neuroscience 1997;8:893–926. [22] Braun AR, Balkins TJ, Wesenten NJ, et al. Regional cerebral blood flow throughout the sleep–wake cycle – an (H2O)-O-15 PET study. Brain 1997;120:1173–97. [23] Hobson JA. The dreaming brain. New York: Basic Books, Inc.; 1988. [24] Szezko PR, Goldberg E, Gunduz-Bruce H, et al. Smaller anterior hippocampal formation volume in antipsychoticnaive patients with first-episode schizophrenia. Am J Psychiatry 2003;160:2190–7. [25] Lipska BK. Using animal models to test a neurodevelopmental hypothesis of schizophrenia. J Psychiatry Neurosci 2004;29:282–6. [26] Hill K, Mann L, Laws KR, Stephenson CM, Nimmo-Smith I, McKenna PJ. Hypofrontality in schizophrenia: a metaanalysis of functional imaging studies. Acta Psychiatr Scand 2004;110:243–56. [27] Laurie SM, Buechel C, Whalley HC, Frith CD, Friston KJ, Johnstone EC. Reduced frontotemporal connectivity in schizophrenia associated with auditory hallucinations. Biol Psychiatry 2002;15:1008–11. [28] Spanos NP, Coe WC. A social–psychological approach to hypnosis. In: Fromm E, Nash MR, editors. Contemporary hypnosis research. New York: The Guilford Press; 1992. p. 102–30. [29] Raz A, Fan J, Posner MI. Hypnotic suggestion reduces conflict in the human brain. Proc Natl Acad Sci USA 2005;102:9978–83. [30] Derbyshire SW, Whalley MG, Stenger VA, Oakley DA. Cerebral activation during hypnotically induced and imagined pain. Neuroimage 2004;23:392–401. [31] Farvoldeu P, Woody EZ. Hypnosis, memory, and frontal executive functioning. Int J Clin Exp Hypn 2004;52: 3–26.
http://intl.elsevierhealth.com/journals/mehy
The conversational brain: Fronto-hippocampal interaction and disconnection J. David Johnson 2021 Jolly Road, Okemos, MI 48864, United States Received 6 April 2006; accepted 10 April 2006
Summary The paper promotes the view that the alert brain alternates between operating in an action mode, based on frontal lobe function, and a receptive mode, involving cholinergic system activity. Their alternation forms a conversation with the environment. It is hypothesized that competition between the modes centers on control over excitability of neurons in the CA1 field of the hippocampus. Increased excitability enhances the flow of hippocampal output through the subiculum resulting in support for frontal lobe function and the action mode. Decreased excitability, on the other hand, reduces this output and that support, leading to a disconnection between frontal lobes and hippocampus. At the same time, correlated cholinergic activity enhances receptive mode processes, indicated by the occurrence of the hippocampal theta rhythm. It is suggested that the hypothesis provides a conceptual framework for considering various phenomena including REM sleep, schizophrenia, and hypnosis. In REM sleep the receptive mode remains dominant as cholinergic activity supports the hippocampal integration of experience into a composite view of reality. In schizophrenia, the action and receptive modes are not properly coordinated because of a dysfunction in anterior hippocampal output. And hypnosis might be seen as a process in which conditions and suggestions are able to induce in some people a prolonged occurrence of the receptive mode allowing a normal view of reality to be altered. c 2006 Elsevier Ltd. All rights reserved.
Introduction The frontal lobes are generally seen as providing the basis for the brain’s executive ability [1]. Their function underlies the initiation, control and maintenance of planned, goal-oriented behavior. Furthermore, they are responsible for mechanisms that protect this behavior, possibly by suppressing activity that would interfere with its realization [2]. However, it seems that there are conditions where the expression of executive control is unE-mail address: [email protected].
needed and even disruptive. These fall into two opposite categories. On the one hand, when behavior is highly practiced and the environment is predictable, it may be best to allow the practiced behavior to prevail over ‘‘second guessing’’ by executive management. On the other hand, when the environment suddenly becomes uncertain or novel experiences appear, it may be adaptive to suspend ongoing planned behavior because such behavior may no longer be appropriate under the new conditions. The need to adjust to changing conditions suggests that there exists a level of control above the executive power of the frontal lobes that is
0306-9877/$ - see front matter c 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2006.04.031
760 able to undermine or interrupt behavior under their direction. It would enable the brain to switch from an executive or action mode to a receptive mode, a mode favoring investigation. As the individual encountered novel experiences and developed ways to deal with them, these modes would alternate back and forth. In effect, interaction with the environment would be conversational, alternating between receptive (e.g., listening) and action (e.g., speaking). This paper hypothesizes that the ability to control the alternation between action and receptive modes rests with the hippocampus. Hippocampal output via the subiculum is seen as a facilitator of executive functions of the frontal lobes thus supporting the action mode. On the other hand, restriction of that output results in an attenuation of those functions and occurs together with a facilitation of the receptive mode. During the action mode, the frontal lobes and hippocampus interact and work together in an integrated fashion, but in the receptive mode there is a disconnection between frontal lobes and hippocampus. The hypothesis is presented as a framework for interpreting the relation between executive and receptive brain functions in various situations. These include the normal waking alert state, REM sleep, schizophrenia, and hypnosis.
Hippocampal theta and the receptive mode The hypothesis is based on the idea that there exist certain brain functions that provide executive control and others that encourage the reception and processing of new information. It is assumed that executive control is a product of frontal lobe function. Receptive processes will be associated with the activity of cholinergic systems. There is strong evidence implicating the cholinergic systems with the creation of conditions that favor the reception of information. Studies of the brainstem cholinergic system have shown that its activity facilitates the relay of sensory signals through the thalamus [3]. Projections to the cortex from the basal forebrain cholinergic system appear to enhance the receptivity of cortical neurons [4]. The medial septal nucleus is part of the basal forebrain system and it projects to the hippocampus. The activity of this projection is widely seen as an important cause of the hippocampal theta rhythm [5]. Numerous studies have been directed at understanding the significance of this rhythm. Many of
Johnson these observed the behavior that accompanies the appearance of hippocampal theta. In a review of such studies in several species, Miller [5] concluded that theta was most likely to appear at times when it was important for the organism to be gathering information about its environment. This was illustrated in a study by Kemp and Kaada with cats [6]. They observed that during waking, maximal hippocampal theta activity was associated with ‘‘fixed staring and visual searching behavior’’. These general observational findings are consistent with the idea that hippocampal theta signals the occurrence of a receptive state. More specific evidence for this can be seen in a study by Tesche and Karhu [7] of human hippocampal activity during a working memory task. They observed that the presentation of the to be remembered stimuli invoked theta, and after several trials theta onset in the right hippocampus began to anticipate the appearance of these stimuli. A burst of hippocampal theta also occurred in response to the probe event that provided the subject the information to properly categorize the stimulus set. These results suggest that theta occurs when the individual prepares to receive information and when the information is being processed. We would associate both conditions with the receptive mode.
Mechanisms underlying frontohippocampal interaction So far it has been proposed that there exists an action mode based on frontal lobe function and a receptive mode based on cholinergic system activity and associated particularly with hippocampal theta activity. The hypothesis includes the idea that these modes compete with each other for dominance. There is evidence to suggest that the competition centers on the control over hippocampal output flowing from the CA1 field and subiculum. It will be argued that restricting this output favors the receptive mode whereas its facilitation benefits the action mode. The evidence begins with experiments showing that during the occurrence of hippocampal theta, output projecting through the subiculum from CA1 is reduced. Winson [8] studied the flow of activity through the hippocampus during several states including theta during waking. He found that at CA1, the last stage in this flow, activity was consistently diminished during theta either in waking or in REM sleep. Other researchers found similar suppression of CA1 spikes in experiments that specifically invoked theta in several ways [9,10]. The
The conversational brain: Fronto-hippocampal interaction and disconnection suppression of CA1 excitability during theta has been shown to originate in reticular formation activation reaching the hippocampus via supramammillary and septal regions [11]. But there is also evidence of an opposite influence on CA1 excitability. It has been shown that input to the hippocampus from the thalamic nucleus reuniens can increase CA1 excitability [12]. McKenna and Vertes [13] observed that the prefrontal cortex provides a major input to the nucleus reuniens and suggested that the nucleus supports an important route of communication between prefrontal cortex and the hippocampus. Thus frontal lobe activity could result in a facilitation of the flow of hippocampal output through CA1 and the subiculum. The significance of a competition for control over hippocampal output can be seen in studies of the effects of that output on target structures. Its effect on the nucleus accumbens is particularly interesting. Grace and colleagues have studied this effect in the context of a circuit that goes from accumbens to ventral pallidum to mediodorsal thalamic nucleus to prefrontal cortex and back to accumbens [14]. They observed that hippocampal output to the accumbens enabled flow through this circuit by allowing activation of accumbens cells by prefrontal cortex to be relayed on to the ventral pallidum. In discussing these results they suggested that the hippocampus is therefore able to influence the level of activity in the prefrontal cortex. By this means, frontal lobe function could be attenuated when CA1 excitability is limited during theta, but could be enhanced when activity originating in the prefrontal cortex via nucleus reuniens leads to increased CA1 excitability and hippocampal output. The circuit implicated in the model of O’Donnell and Grace [14] could be characterized as providing motivational support for frontal lobe function [15]. The parahippocampal gyrus provides another basis for interaction between hippocampus and prefrontal cortex that may influence the specific nature of frontal lobe activity. The gyrus receives output from the subiculum and relays it on to many cortical areas including prefrontal cortex [16]. Previously it was suggested that the wide spread projection mediated by the parahippocampal gyrus supports the imposition of a contextual influence upon target structures [17]. If that is true, control of hippocampal output at CA1 might not only determine the motivation for pursuing goal-oriented behavior but might also have an effect on processes shaping the direction of this behavior. In short, it appears that during hippocampal theta and the occurrence of the receptive mode, hip-
761
pocampal output is restricted at CA1. This creates a disconnection between hippocampus and frontal lobes that may lead to an attenuation of frontal lobe function and suspension of the action mode. On the other hand, when frontal lobe activity increases enabling the action mode, the frontal lobes and hippocampus are able to interact. The presence of the action mode may be reinforced as increased CA1 excitability allows hippocampal output to support frontal lobe function.
Interaction and disconnection during normal waking Previous papers presented a view of hippocampal function which centered on the idea that the hippocampus alternates between a theta mode and a non-theta mode [17–19]. During the theta mode, input to the hippocampus is unrestricted but output through the subiculum is restricted. While in this mode, a variety of information from throughout the brain becomes integrated in the hippocampus. This results in the formation of context memory which is a composite picture of the reliable aspects of current experience. This memory serves as an integrated set of expectations about the environment, a worldview. By comparison, during the non-theta mode, input to the hippocampus is restricted while output through the subiculum is unrestricted. The pattern of this output is based on the context memory formed during the theta mode and possibly becomes copied into the parahippocampal gyrus. The output is seen as having two major functions. First, it creates throughout the neocortex a scaffold on which the patterns of specific episodes may become attached. These associations provide the basis for retrieval of the memory of episodes. Secondly, the output creates a widespread contextual influence. The function of that influence is to cause perception, emotion and behavior to be consistent with the underlying context memory. The influence was described as providing a worldview and reality orientation. Orientation in space was linked to output from the dorsal hippocampus whereas emotional orientation was attributed to output from the anterior hippocampus. The concepts of receptive and action modes described here correspond to and are extensions of the theta and non-theta modes described previously. When the receptive mode is present during normal waking, it represents a naturally occurring disconnection between hippocampal and frontal lobe functions. It is expected that this sort of
762 disconnection would be most prominent when individuals are young or later when they are placed in a novel environment, i.e., when there is a maximum amount of new information that should be acquired in order to adapt. But as individuals mature and conditions become familiar, much less time would be spent in the receptive mode and the action mode would become increasingly dominant during alert waking, especially in species with highly developed frontal lobes. This may explain why it has been difficult to observe hippocampal theta in adult primates [20], and when it is seen, it occurs in brief bursts [7]. In addition to degree of experience and species differences, other factors could determine the extent to which the action or receptive mode is dominant. For instance, if the novel event were immediately perceived as threatening, individuals may forgo extensive investigation and quickly begin deriving plans for action. But if the novel event were merely a curiosity and promoted relaxation, they might remain longer in the receptive mode. Furthermore, individuals might naturally vary in the extent to which one or the other mode prevailed. Some might maintain executive control almost continuously whereas others might more readily suspend this control.
Other cases of interaction and disconnection Previously the phenomena of REM sleep and of schizophrenia were discussed in terms of hippocampal function [18,19]. Here those discussions will be extended to consider the phenomena in terms of the concepts of action and receptive modes. Then the discussion will turn to a consideration of the nature of hypnosis from the perspective of this conceptual framework.
REM sleep Periods of REM sleep coincide with the appearance of hippocampal theta [21]. Therefore we would expect that REM sleep is a time in which the receptive mode remains dominant and frontal lobes are disconnected from the hippocampus. Evidence of the disconnection can be found in measures of brain activity during REM. Braun et al. [22] observed a marked reduction in frontal lobe activity during REM when compared either to waking or slow wave sleep. At the same time, the hippocampus and parahippocampal cortices showed increased activation.
Johnson REM sleep is also commonly associated with the occurrence of vivid dreaming [23]. A reflection on the nature of dreams suggests support for the idea that they occur when the action mode has been replaced by the receptive mode. Dreams lack the planning and executive control that is characteristic of the action mode. Instead, they often seem to be series of experiences that we passively observe or respond to reflexively, as would be expected when one is in the receptive mode. According to the present view, REM sleep is essentially a period of input processing where the input is a variety of information flowing into the hippocampus. It has been hypothesized that during REM sleep conditions are ideal for developing a composite memory of current experience [18]. The reduction in frontal lobe activity during this time was said to allow a free flow of information, free of the bias of executive control.
Schizophrenia Over the years various hypotheses have been offered to explain the cause or causes of schizophrenia. Recent measures of brain activity in schizophrenics suggest that dysfunction within the medial temporal lobe, the anterior hippocampus in particular, may be an underlying cause of this disease [24]. This idea has gained support from studies showing that hippocampal lesions in neonatal rats can cause schizophrenia-like symptoms when they mature [25]. Dysfunction of the anterior hippocampus would create an unnatural disconnection between the hippocampus and the targets of its output. Previously, it was argued that fundamental schizophrenic symptoms, such as inappropriate affect and splitting of associations, were caused by a failure of anterior hippocampal output to control its target structures [19]. Inappropriate affect was the result of a lack of a contextual influence upon emotional processes. The splitting of associations was caused by a failure of this output to support frontal lobe control over the flow of associations. When viewed in terms of the idea of competing action and receptive modes, schizophrenia could be seen as a condition in which these modes are not properly coordinated. Dysfunction of the anterior hippocampus could create a lasting disconnection between hippocampal and frontal lobe activity. This may account for the finding of reduced frontal lobe activity in some schizophrenics [26]. O’Donnell and Grace [14] suggested that their hippocampal gating model might explain the cause of this hypofrontality. However, not all schizophrenics show decreased frontal lobe activity. For
The conversational brain: Fronto-hippocampal interaction and disconnection example, Laurie et al. [27] observed brain activity in schizophrenics and normals and reported both absolute levels and correlations between frontal and temporal lobe activity. They found that levels of frontal lobe activity did not differ between normals and patients. But activities of frontal and temporal lobes were significantly less correlated in schizophrenics. These results suggest that frontal lobe activity can achieve a normal level without the support of anterior hippocampal output, but that this output may be necessary for proper coordination of the functions of frontal and temporal lobes including the control of competition between action and receptive processes.
Hypnosis The term hypnosis refers to a condition in which it is said that sensations, perceptions, emotions, and behavior can be altered by suggestion. However, the existence of such a state remains controversial. The present paper does not take a position on the controversy. But it does suggest that the concepts of action and receptive modes provide a useful framework for discussing hypnosis and point to a way to test for its existence. Critics of the existence of hypnosis maintain that the subjects of a hypnotic session are behaving in a goal directed manner and are trying to satisfy the expectations and wishes of the hypnotist [28]. In terms of the current framework, the subjects remain predominantly in the action mode during and after a session. Frontal lobe functions remain in control throughout the experience. However, it is possible to imagine a different scenario. Conceivably, for some people the stimuli and instructions presented during hypnotic induction might elicit the receptive mode such that it becomes dominant and sustained. If that were to occur, we would expect that persons would accept their experience uncritically (as happens during REM sleep dreaming) due to an attenuation of frontal lobe function. And the information they receive in this state, including hypnotic suggestions, could be incorporated in the worldview that is formed when in the receptive/theta mode [18,19]. As a result, persons could emerge from a session with an altered view of reality that is enforced when the action mode regains control. To establish that the events in this scenario could actually occur, one needs evidence from measures of brain function. It is important to distinguish two kinds of evidence. The first would be measures of the effects of hypnotic suggestions. These results could be as varied as the suggestions
763
can be. For instance, hypnotic suggestions intended to reduce conflict were shown to decrease activity in the anterior cingulate cortex [29] whereas suggestions that pain will be experienced lead to increased activity in this cortex [30]. The second kind of evidence would be measures of underlying conditions that allow a person to be hypnotized. The conceptual framework promoted here predicts the results one would expect if hypnosis exists. There should be evidence associating hypnotizability with decreased frontal lobe function (which would weaken the control of the action mode) and increased occurrence of hippocampal theta, especially during hypnotic induction (signaling the dominance of the receptive mode). A reduction in measures of frontal lobe function in highly hypnotizable subjects has already been reported [31]. It remains to be seen whether these subjects also show an above average increase in hippocampal theta during a hypnotic session.
Conclusion This paper promotes the idea that brain function alternates between a receptive mode and an action mode. The receptive mode favors the gathering of information whereas the action mode supports the generation of goal-oriented behavior. The action mode relies on the activity of recently evolved neocortex of the frontal lobes while the receptive mode is based on much older cholinergic systems. From an evolutionary perspective their interaction is a competition between newer and older brain functions. If the difficulty to observe hippocampal theta in primates is any guide, it appears that the action mode in mammals is becoming the stronger competitor in those with larger frontal lobes [20]. However, there will always be a need to gather and process information from our environment. Executive control would be acting blindly if receptive mode processes were entirely overridden by the action mode. So we would expect the receptive mode to remain competitive. If it were shown that hypnotizability really exists and is a trait that allows the receptive mode to be easily invoked, we could speculate that this trait is adaptive in humans in as much as it enables the receptive mode to remain competitive in the face of a highly assertive action mode.
References [1] Goldberg E. The executive brain. Oxford: Oxford University Press; 2001.
764 [2] Fuster JM. The prefrontal cortex: anatomy, physiology and neuropsychology of the frontal lobes. 3rd ed. Philadelphia: Lippincott-Raven; 1997. [3] Steriade M, Buzsaki G. Parallel activation of thalamic and cortical neurons by brainstem and basal forebrain cholinergic systems. In: Steriade M, Biesold D, editors. Brain cholinergic systems. Oxford: Oxford University Press; 1990. p. 1–62. [4] McCormick DA. Cellular mechanisms of cholinergic control of neocortical and thalamic neuronal excitability. In: Steriade M, Biesold D, editors. Brain cholinergic systems. Oxford: Oxford University Press; 1990. p. 236–64. [5] Miller R. Cortico-hippocampal interplay and the representation of contexts. Berlin: Springer-Verlag; 1991. p. 83–85. [6] Kemp IR, Kaada BR. The relation of hippocampal theta activity to arousal, attentive behaviour and somato-motor movements in unrestrained cats. Brain Res 1975;95:323–434. [7] Tesche CD, Karhu J. Theta oscillations index human hippocampal activation during a working memory task. Proc Natl Acad Sci USA 2000;97:919–24. [8] Winson J. Neuronal transmission through the hippocampus: dependence on behavioral state. In: Reinoso-Suarez F, Ajmone-Marsan C, editors. Cortical integration. New York: Raven Press; 1984. p. 131–46. [9] Herreras O, Solis JM, Munoz MD, Martin del Rio R, Lerma J. Sensory synapsis modulation of hippocampal transmission. I. Opposite effects on CA1 and dentate gyrus. Brain Res 1988;461:290–302. [10] Vinogradova OS, Brazhnik ES, Kitchigina VF, Stafekhina VS. Acetylcholine, theta-rhythm and activity of hippocampal neurons in the rabbit – IV. Sensory stimulation. Neuroscience 1993;53:993–1007. [11] Jiang F, Khanna S. Reticular stimulation evokes suppression of CA1 synaptic responses and generation of theta through separate mechanisms. Eur J Neurosci 2004;19:295–308. [12] Bertram EH, Zhang DX. Thalamic excitation of hippocampal CA1 neurons: a comparison with the effects of CA3 stimulation. Neuroscience 1999;92:15–26. [13] McKenna JT, Vertes RP. Afferent projections to nucleus reuniens of the thalamus. J Comp Neurol 2004;480:115–42. [14] O’Donnell P, Grace AA. Synaptic interactions among excitatory afferents to nucleus accumbens neurons: hippocampal gating of prefrontal cortical input. J Neurosci 1995;15:3622–39. [15] Chow TW, Cummings JL. Frontal-subcortical circuits. In: Miller BL, Cummings JL, editors. The human frontal lobes: functions and disorders. New York: The Guilford Press; 1999. p. 3–26.
Johnson [16] VanHoesen GW. The parahippocampal gyrus: new observations regarding its cortical connections in the monkey. Trends Neurosci 1982;5:345–50. [17] Johnson JD. Episodic memory and the hippocampus: another view. Med Hypoth 2004;63:963–7. [18] Johnson JD. REM sleep and the development of context memory. Med Hypoth 2005;64:499–504. [19] Johnson JD. Dysfunction of the anterior hippocampus: the cause of fundamental schizophrenic symptoms? Med Hypoth 2005;65:55–60. [20] Stewart M, Fox SE. Hippocampal theta activity in monkeys. Brain Res 1991;538:59–63. [21] Vertes RP, Kocsis B. Brainstem-diencephalo-septohippocampal systems controlling the theta rhythm of the hippocampus. Neuroscience 1997;8:893–926. [22] Braun AR, Balkins TJ, Wesenten NJ, et al. Regional cerebral blood flow throughout the sleep–wake cycle – an (H2O)-O-15 PET study. Brain 1997;120:1173–97. [23] Hobson JA. The dreaming brain. New York: Basic Books, Inc.; 1988. [24] Szezko PR, Goldberg E, Gunduz-Bruce H, et al. Smaller anterior hippocampal formation volume in antipsychoticnaive patients with first-episode schizophrenia. Am J Psychiatry 2003;160:2190–7. [25] Lipska BK. Using animal models to test a neurodevelopmental hypothesis of schizophrenia. J Psychiatry Neurosci 2004;29:282–6. [26] Hill K, Mann L, Laws KR, Stephenson CM, Nimmo-Smith I, McKenna PJ. Hypofrontality in schizophrenia: a metaanalysis of functional imaging studies. Acta Psychiatr Scand 2004;110:243–56. [27] Laurie SM, Buechel C, Whalley HC, Frith CD, Friston KJ, Johnstone EC. Reduced frontotemporal connectivity in schizophrenia associated with auditory hallucinations. Biol Psychiatry 2002;15:1008–11. [28] Spanos NP, Coe WC. A social–psychological approach to hypnosis. In: Fromm E, Nash MR, editors. Contemporary hypnosis research. New York: The Guilford Press; 1992. p. 102–30. [29] Raz A, Fan J, Posner MI. Hypnotic suggestion reduces conflict in the human brain. Proc Natl Acad Sci USA 2005;102:9978–83. [30] Derbyshire SW, Whalley MG, Stenger VA, Oakley DA. Cerebral activation during hypnotically induced and imagined pain. Neuroimage 2004;23:392–401. [31] Farvoldeu P, Woody EZ. Hypnosis, memory, and frontal executive functioning. Int J Clin Exp Hypn 2004;52: 3–26.