Secondary Logo

Journal Logo


How an Understanding of the Function of the Locus Coeruleus Led to Use of Dexmedetomidine to Treat Agitation in Bipolar Disorder: Example of Rational Development of Psychiatric Medications

Preskorn, Sheldon H. MD

Author Information
Journal of Psychiatric Practice: May 2022 - Volume 28 - Issue 3 - p 227-233
doi: 10.1097/PRA.0000000000000634
  • Free


I was fortunate to have been the lead author on the lead article published in the February 22, 2022, issue of the Journal of the American Medical Association (JAMA).1 This paper detailed the efficacy and tolerability/safety of dexmedetomidine, a full and selective alpha-2 adrenergic agonist, for the treatment of agitation in patients with bipolar disorder. Dexmedetomidine is a full alpha-2 adrenergic agonist, meaning it has intrinsic activity at this receptor comparable to that of the endogenous neurotransmitter, norepinephrine (NE); it is also a selective agonist, meaning its affinity for the alpha adrenergic-2 receptor is ≥10 times greater than for any other known receptor. I was fortunate because, ∼50 years earlier, numerous colleagues and I were involved in determining the functional role of the locus coeruleus (LC) in the brain and how it could mediate a wide range of psychopathologic states in man. In this column, I present a longitudinal view of how this knowledge moved from initial discovery to this most recent research published in JAMA (ie, going from A to Z, with A being the first step in understanding the basic science of the LC and Z being the latest step, the translation of that science into a potential treatment for agitation in a psychiatric illness).


Anatomy is often the first step in understanding pathophysiology, and that was the case with the LC (which is also referred to as the CAS). The LC lies under the floor of the fourth cerebral ventricle which provides a first clue to its function (Fig. 1). It is comprised of 20,000 to 50,000 neurons, which produce the neurotransmitter NE (also called noradrenaline, hence the adjective, noradrenergic). These neurons are pigmented and densely packed together. Hence, the LC can be seen by visual inspection of the brain and was first identified by Félix Vicq-d’Azyr in 1784.2,3

Diagram of the locus coeruleus. The locus coeruleus is a tightly packed cluster of norepinephrine neurons sitting under the fourth ventricle of the brain. Some of the fibers from these neurons project down the spinal cord so that the locus coeruleus can modulate and coordinate the cardiovascular and respiratory systems. Others fibers project upwards to innervate most, if not all, regions of the brain. Some fibers project into the fourth ventricle and react to changes in acid-base balance in the brain. This anatomy is, in part, what underlies fight or flight reactions, panic attacks, and agitation, to name a few conditions mediated by these brainstem nuclei. Figure was obtained from Shutterstock/Vasilisa Tsoy.

However, elucidating the anatomy of the CAS required the development of tools to map its fiber tracts. In 1977, I was a senior resident in psychiatry at Washington University in St Louis, Missouri, and I was being mentored by a professor in that department, Boyd Hartman, MD. A few years earlier, Dr. Hartman had developed a fluorescent-tagged antibody for the enzyme dopamine-beta-hydroxylase which converts dopamine into NE.

Using this antibody tool, Dr. Hartman and his colleague, Dr. Swanson, described the anatomy of the LC, including: (a) that the fibers arising out of some of its neurons penetrate into the cerebrospinal fluid of the fourth ventricle and (b) that other fibers from these neurons abut against the cerebral microvasculature appearing to innervate them.4 In relatively short order, receptors for NE were identified on the cerebral microvasculature, further supporting the conclusion that the LC innervated this vasculature.5 These 2 observations suggested that the LC could connect the monitoring of the acid-base balance in the cerebrospinal fluid to both blood-brain barrier (BBB) permeability to small molecules (including water!) and the regional cerebral blood flow (CBF) in the brain. The question then was whether this hypothesis could be tested physiologically and whether it had relevance to any psychiatric disorders or states.


While the research described above was ongoing, I began working in the Hartman laboratory on the pharmacology of the LC and its potential role in regulating BBB permeability to water and other small molecules (principally ethanol) and CBF in response to acid-base disturbances in the brain. Several techniques were developed to study these phenomena in rats and primates.6–9 These techniques were based on using radioisotopes of small molecules (eg, water and ethanol) to measure regional CBF [regional here refers to relatively large structures, particularly in rats (eg, cerebral cortex vs. thalamus) rather than the much more precise regional differences now possible]. Parenthetically, fundamentally similar approaches are employed today, such as using positron-emitting radioisotopes [positron emission tomography (PET)] to do such measurements in humans with much greater regional specificity for both research and clinical purposes.

One of the first findings from the use of these techniques was that water and ethanol are not completely free in their ability to cross the BBB but instead have an extraction fraction which is close to but not 1 (ie, not completely permeable). Using these techniques, 2 potential functions of the LC could be measured: (1) BBB permeability to water and ethanol and (2) CBF. These 2 measurements and their implications form the basis of most of the rest of the discussion and are referred to as effects.

For historical context, tricyclic antidepressants (TCAs) were the most widely used type of antidepressant from the 1960s through the 1980s. In addition, they were known to work on the presynaptic NE transporter and serotonin transporter, respectively. Parenthetically, these transporters have also been called “pumps,” particularly in the clinical literature. I wanted to study the effect of these drugs for 2 reasons: their mechanism affected the LC, and they were clinically relevant.

The following findings came out of this work:

  • Pretreatment with TCAs could alter BBB permeability and CBF both acutely and chronically, mimicking the clinical use of the drugs.10–13
  • These effects occurred in primate as well as rodent brains, and they also altered regional brain metabolic rate in primates.14
  • These effects in both rats and primates occurred at plasma concentrations of TCAs that were comparable to therapeutic plasma levels in patients with major depression being treated with TCAs.15 Thus, these physiological changes were likely occurring in patients being treated for psychiatric illnesses with these medications.
  • TCAs, particularly tertiary amine TCAs such as amitriptyline, have effects on multiple targets at concentrations near or even lower than what is needed to affect the NE transporter. Hence, it was possible that one of those mechanisms contributed or accounted for the effect. Hence, we tested drugs which—as far as we knew at that time—only affected histamine receptors, muscarinic acetylcholine receptors, or alpha-1 adrenergic receptors; however, individually, none of these selective drugs caused the same changes in BBB permeability to water or ethanol and/or in CBF.
  • In contrast to these negative findings, the changes in CBF and BBB permeability induced by TCAs were dependent on an intact LC, given that they were abolished by treatments that either irreversibly destroyed (or ablated) the LC or temporarily blocked the effects of the LC by the use of drugs such as tetrabenazine which depletes biogenic amines or phenobenzamine which blocks alpha-1 adrenergic receptors. In addition, the effects of TCAs on BBB permeability and CBF were not affected by destroying central serotonin or dopamine neurons.16–18 Those studies supported the hypothesis that these drugs were working through the LC rather than acting directly on CBF or BBB permeability.
  • Changes in BBB permeability and CBF were also observed in pharmacological and some animal models of human major depression, further suggesting that these changes could be relevant to major depression in humans.19–21
  • The changes in CBF and BBB permeability also changed the volume of extracellular space. Changes in this parameter altered the sodium: potassium balance across the neuronal membrane, which suggests a means by which the LC could modulate the sensitivity of other neurons to afferent inputs in regions innervated by the LC, as has been described by other researchers.22

For these initial findings concerning the effect of TCAs on cerebral fluid dynamics, I was awarded the A.E. Bennett Neuropsychiatric Research Foundation Award from the Society of Biological Psychiatry in 1978,23 and, for subsequent research, I was funded by a Research Scientist Development Award from the National Institute of Mental Health.


At the time this work was ongoing, studies reported from the laboratory of TH Swensson of Sweden established that the firing of LC neurons and CBF were directly regulated by the concentration of CO2 in the blood.24 As CO2 increased, LC neuronal firing and CBF increased. That observation suggested that the LC was a central mechanism in sensing and guarding against hypercapnia and hypoxia by serving essentially as a central fire alarm system when such states occurred and activating the animal to escape the dangerous environment.

Using this knowledge, we demonstrated that increasing the firing of the LC by increasing the concentration of inhaled CO2 enhanced the effect of TCAs, increasing both BBB permeability to water and ethanol and CBF. These findings further supported the conclusion that TCAs were working on the LC to change BBB permeability and CBF in response to a physiological challenge: hypercapnia or hypocapnia.22

An extension of this research led me, along with multiple other colleagues, most notably Wayne C. Drevets, MD, to conduct the first study that defined the functional neuroanatomy of major depression.25 The findings of this study have been cited >1500 times in the world’s medical literature and have been replicated many times during the past couple of decades.


As this work was progressing, I became aware of a parallel line of research examining acid-base balance and panic disorder. In 1951, Cohen and White26 reported that patients with panic disorder frequently had acid-base disturbances when they had panic attacks. In 1967, Pitts and McClure27 demonstrated that infusion of substances such as sodium lactate, or later CO2 inhalation, could produce panic attacks in individuals who had spontaneous panic attacks at concentrations lower than would be needed to produce panic attacks in normal controls.

These observations suggested to me that these findings, combined with the other work described above, implicated the LC as having a role in the pathobiology underlying panic disorder. In the early 1980s, I had the occasion to meet one-on-one with Donald Klein when he visited the Kansas University School of Medicine, and I discussed this line of work with him, given his interest in panic disorder. In 1993, Klein28 published his seminal work on this hypothesis. Readers interested in this area of research are referred to a review article by Wemmie.29

In addition to these findings, the LC has been implicated in the arousal system in the brain from sleep to wakefulness.30,31 The LC thus plays a fundamental role in regulating the degree of arousal of the brain—from sleep to arousal to agitation to panic, with these states occurring along a biologically mediated continuum.

Returning to its anatomy (Fig. 1), the LC has fibers that extend (1) caudally to regulate cardiac output and peripheral arterial resistance and (2) rostrally to arouse or hyperarouse the brain. The effect of the first set of fibers accounts for the symptoms of pounding heart, cold and clammy extremities, and hyperventilation seen in panic disorder, the flight or fight response, and agitation. The effect of the second set of fibers accounts for hyperarousal, anxiety, fear, and agitation.


The research reviewed above supports the hypothesis that a full alpha adrenergic-2 agonist such as dexmedetomidine should be able to treat a hyperaroused state such as agitation. That hypothesis was tested and supported by the study published in JAMA mentioned at the beginning of this column.1

To understand the hypothesis that drove this study, one must understand the role of the alpha-2 adrenergic receptor on the firing of the LC. This process is illustrated in Figure 2, which has been called the standard National Institute of Health model of biogenic amine synaptic transmission. The presynaptic neuron has a receptor for the neurotransmitter released by that specific biogenic amine neuron (ie, an autoreceptor because it responds to the neurotransmitter released by that neuron).

Components of a biogenic amine synapse. This figure illustrates many of the components of a biogenic amine (ie, norepinephrine, dopamine, and serotonin) synapse. With regard to activity discussed in this article, the neuron produces and releases norepinephrine and the critical element is the alpha-2A autoreceptor on the terminal projections. These receptors also exist on the cell body but that is not shown in this illustration. The cell self-regulates its release of norepinephrine via this autoreceptor, which decreases the release of norepinephrine when the norepinephrine concentration in the synapse becomes too high, especially in acute situations. If this mechanism does not work adequately in a given situation (eg, escalating fear or agitation or a stimulus causing the neurons to fire), alpha-2A agonists can be administered that work on this mechanism to decrease the release of norepinephrine. Reprinted with permission from Richelson.32

In the case of the LC neurons, the alpha-2 adrenergic receptor is the autoreceptor that has the function of decreasing firing and the release of NE when the concentration of NE gets too high in the synapse, as can happen in pathologic states such as fight or flight reactions, panic attacks, anxiety states, and agitation to name a few. Parenthetically, 3 different forms of the alpha-2 adrenergic receptor called A, B, and C are differentially expressed regionally in the brain, with alpha-2A being preferentially expressed on LC neurons.

Alpha-2 adrenergic drugs are characterized as agonists, partial agonists, or antagonists. An extensive discussion of these drugs is beyond the scope of this paper. Instead, the focus will be on full or partial alpha-2A adrenergic agonists. In evaluating any type of drug, one considers 3 major characteristics: (a) what is the activity of the drug at its receptor(s) (ie, full or partial agonist, antagonist, full or partial inverse agonist)? (b) does the drug have high or low binding affinity for its receptor(s)? and (c) is the drug selective for one specific receptor or does it affect other types of receptors at similar concentrations?

Table 1 shows the first 2 characteristics of several currently available alpha-2 adrenergic receptor agonists: (a) affinity for this receptor and (b) intrinsic activity on the receptor. Intrinsic activity is expressed relative to the activity of the endogenous neurotransmitter, NE. As can be readily seen in this table, dexmedetomidine was chosen for the treatment of agitation because of its high affinity for the alpha-2A receptor (ie, EC50 is the inverse of affinity, meaning the lower the concentration of the drug needed to affect the receptor, the higher its affinity for that receptor) and its high intrinsic action on the receptor (ie, % of maximal activity relative to the endogenous neurotransmitter). Parenthetically, these 4 drugs show a rank order comparable to what is shown in the table for the 2A receptor in their affinity and intrinsic activity at the 2B and 2C variants of the receptor. Dexmedetomidine is also more selective than the other agents shown in the table in terms of not binding to other receptors to any appreciable extent, but that discussion is beyond the scope of this paper.

TABLE 1 - Comparison of the Affinity and Intrinsic Activity of Marketed Alpha-2A Adrenergic Agonists33,34
Drugs* EC50 (nM) Maximal Activity (%)
Clonidine 25 76
Dexmedetomidine 4 89
Guanfacine 69 78
Lofexidine 17 40
*Drug listed by generic name in alphabetical order.
EC50 refers to half maximal effective concentration: the concentration required to obtain a 50% effect. The EC50 is a measure of the affinity of the drug for a receptor: the lower the number, the higher the affinity for the receptor.
A measure of the agonist properties of the drug relative to norepinephrine, the endogenous neurotransmitter.


This column outlines the history underlying our understanding of the role of the LC in the regulation of arousal from sleep to wakefulness to hypervigilance to anxiety to flight or fight to panic attacks to agitation. The anatomy of the LC is consistent with its role in the response to stimuli that increase arousal leading to the fight or flight reaction. The LC is located just below the floor of the fourth ventricle and sends fibers into the cerebrospinal fluid to sense the acid-base balance. When that balance is disturbed by stimuli such as hypercapnia and/or hypoxia, the LC neurons increase their firing to increase sympathetic tone in the brain and the periphery. As illustrated in Figure 1, the LC fibers project downward to regulate the respiratory rate and blood pressure by increasing cardiac output and constricting peripheral blood vessels. This process accounts for many of the signs and symptoms seen in a panic attack or a fight or flight reaction and redirects more cardiac output to the brain. The LC also projects upwards to innervate upper brain regions to be more sensitive to afferent inputs and be prepared to act. Panic attacks and agitation can thus occur when the LC responds to perturbations in acid-base balance that would normally not cause a response.

Due to the fundamental role of the LC in both normal brain functioning and responses to perceived threats by increasing its firing rate and activating fight or flight responses, the system must be tightly regulated so that such reactions do not occur without appropriate stimuli (such as happens in panic attacks). This system is hence autoregulated in normal physiological states by the alpha-2 adrenergic receptor which is located on the LC neurons and thus forms an autofeedback loop which, along with other mechanisms (eg, CO2 concentration in the cerebrospinal fluid and blood), regulates the likelihood of the cells firing. As the cells fire, NE increases at the alpha-2 adrenergic receptor on the adrenergic neuron, which reduces the firing rate of the neuron. Dexmedetomidine, as a full and selective alpha-2 adrenergic-receptor agonist, acts like NE to decrease the firing of the LC.

Increased anxiety, fear, and agitation occur along a spectrum of increasing central and peripheral hypersympathetic activity mediated by increasing firing of LC neurons. Dexmedetomidine was thus predicted to be able to reduce this hypersympathetic state and hence agitation by decreasing the firing of LC neurons in a dose-dependent manner. That prediction was successfully tested in agitated patients with bipolar disorder.1

This column also complements multiple previous columns that have discussed how novel central nervous system drugs can be developed based on an understanding of the circuitry underlying the target indications.35–39 These columns also illustrate how specific forms of psychopharmacology can be tied to specific brain mechanisms and circuits, allowing for a more sophisticated way of conceptualizing and utilizing various psychopharmacological therapies.


The work described in this column spanned many years, and many colleagues contributed to it. Those individuals are authors on the references for this column, particularly on the publications on which I am also an author, so I will not list all of their names here. However, I wanted to specifically acknowledge 3 individuals. Boyd Hartman, MD, who was my mentor on these studies, Belinda Preskorn, my wife and supporter through all of this and other work, and Ruth Ross, who has been and is the managing editor of the Journal of Psychiatric Practice and my collaborator on all my columns published in this journal.


1. Preskorn SH, Zeller S, Citrome L, et al. Effect of sublingual dexmedetomidine vs placebo on acute agitation associated with bipolar disorder: a randomized clinical trial. JAMA. 2022;327:727–736.
2. Tubbs RS, Loukas M, Shoja MM, et al. Félix Vicq d’Azyr (1746-1794): early founder of neuroanatomy and royal French physician. Childs Nerv Syst. 2011;27:1031–1034.
3. Swanson LW. Neuroanatomical Terminology: A Lexicon of Classical Origins and Historical Foundations. Oxford, UK: Oxford University Press; 2014.
4. Swanson LW, Hartman BK. The central adrenergic system: an immunofluorescence study of the location of the cell bodies and their efferent connections in the rat utilizing dopamine-beta-hydroxylase as a marker. J Comp Neurol. 1975;163:467–506.
5. Hartman BK, Swanson LW, Kaplan GP, et al. Anatomical evidence for central catecholaminergic innervation of the cerebral microvasculature. J Cereb Blood Flow Metab. 1981;1:S307–S308.
6. Clark HB, Hartman BK, Raichle ME, et al. Measurement of cerebral vascular extraction fractions in the rat using intracarotid injection techniques. Brain Res. 1981;208:311–323.
7. Clark HB, Hartman BK, Raichle ME, et al. An intravenous technique for the measurement of cerebral vascular extraction fraction in the rat. J Cereb Blood Flow Metab. 1982;2:187–196.
8. Irwin GH, Preskorn SH. A dual label radiotracer technique for the simultaneous measurement of cerebral blood flow and the single-transit cerebral extraction of diffusion-limited compounds in rats. Brain Res. 1982;249:23–30.
9. Kent T, Preskorn SH. A modified method for the simultaneous determination of regional single-transit brain extraction of diffusion-limited compounds and cerebral blood flow: utilization of non-invasive measurement of transit time. Brain Res. 1985;335:251–255.
10. Preskorn SH, Hartman BK, Raichle ME, et al. The effect of dibenzazepines (tricyclic antidepressants) on cerebral capillary permeability in the rat in vivo. J Pharmacol Exp Ther. 1980;213:313–320.
11. Preskorn SH, Hartman BK, Raichle ME, et al. Central adrenergic regulation of cerebral microvascular permeability and blood flow: pharmacologic evidence. Adv Exp Med Biol. 1980;131:127–138.
12. Preskorn SH, Hartman BK, Clark HB. Long-term antidepressant treatment: alterations in cerebral capillary permeability. Psychopharmacology (Berl). 1980;70:1–4.
13. Preskorn SH, Irwin GH, Simpson S, et al. Medical therapies for mood disorders alter the blood-brain barrier. Science. 1981;213:469–471.
14. Preskorn SH, Raichle ME, Hartman BK. Antidepressants alter cerebrovascular permeability and metabolic rate in primates. Science. 1982;217:250–252.
15. Glotzbach RK, Preskorn SH. Brain concentrations of tricyclic antidepressants: single-dose kinetics and relationship to plasma concentrations in chronically dosed rats. Psychopharmacology (Berl). 1982;78:25–27.
16. Preskorn SH, Hartman BK, Irwin GH, et al. Role of the central adrenergic system in mediating amitriptyline-induced alteration in the mammalian blood-brain barrier in vivo. J Pharmacol Exp Ther. 1982;223:388–395.
17. Preskorn SH, Irwin G, Simpson S, et al. Modulation of cerebral fluid dynamics: alpha and beta adrenergic mechanisms. J Cereb Blood Flow Metab. 1981;1:S373–S374.
18. Kent TA, Preskorn SH, Solnick JV, et al. Differential effects of central biogenic amines on cerebral blood flow and the blood-brain barrier. J Cereb Blood Flow Metab. 1983;3:S228–S229.
19. Hughes C, Kent T, Campbell J, et al. Cerebral blood flow and cerebrovascular permeability in an inescapable shock (learned helplessness) animal model of depression. Pharmacol Biochem Behav. 1984;21:891–894.
20. Preskorn SH, Kent TA, Glotzbach RK, et al. Cerebromicrocirculatory defects in animal model of depression. Psychopharmacology (Berl). 1984;84:196–199.
21. Kent T, Preskorn SH, Glotzbach RK, et al. Amitriptyline normalizes tetrabenazine-induced changes in cerebral microcirculation. Biol Psychiatry. 1986;21:483–491.
22. Kent T, Nagy G, Oke A, et al. Effect of CO2 on a brain extracellular space marker and evidence of its neuronal modulation. Brain Res. 1985;342:141–144.
23. Preskorn SH, Hartman BK. The effect of tricyclic antidepressants on cerebral fluid dynamics. Biol Psychiatry. 1979;14:235–250.
24. Elam M, Vao T, Thoren PI, et al. Hypercapnia and hypoxia: chemoreceptor mediated control of the locus coeruleus neurons and splanchnic, sympathetic nerves. Brain Res. 1981;222:373–381.
25. Drevets WC, Videen T, Preskorn SH, et al. A functional anatomical study of unipolar depression. J Neurosci. 1992;12:3628–3641.
26. Cohen ME, White PD. Life situations, emotions, and neurocirculatory asthenia (anxiety neurosis, neurasthenia, effort syndrome). Psychosom Med. 1951;13:335–357.
27. Pitts FN, McClure JN. Lactate metabolism in anxiety neurosis. N Engl J Med. 1967;227:1329–1336.
28. Klein DF. False suffocation alarms, spontaneous panics, and related conditions. An integrative hypothesis. Arch Gen Psychiatry. 1993;50:306–317.
29. Wemmie JA. Neurobiology of panic and pH chemosensation in the brain. Dialogues Clin Neurosci. 2011;13:475–483.
30. Spaer CB, Scammel TE, Lu J. Hypothalamic regulation of sleep and circadian rhythms. Nature. 2006;437:1257–1262.
31. Tononi G, Cirreli C Charney DS, Nestler EJ. The neurobiology of sleep. Neurobiology of Mental Illness, Third Edition. Oxford, UK: Oxford University Press; 2009:1370–1386.
32. Richelson E. Pharmacology of antidepressants—characteristics of the ideal drug. Mayo Clin Proc. 1994;69:1069–1081.
33. De Vivo M, Seth S, Sharma S, et al. Dexmedetomidine–highly favorable pharmacokinetic and pharmacological features for a CNS therapeutic drug adaptation to stress. Poster presented at the Annual Meeting of the American College of Neuropsychopharmacology, Orlando, FL, December 7-11, 2019.
    34. Audinot V, Fabry N, Nicolas J-P, et al. Ligand modulation of [35S]GTPgS binding at human a2A, a2B and a2C adrenoceptors. Cell Signal. 2002;14:829–837. Available at: Accessed March 27, 2020.
    35. Preskorn SH. CNS drug development: lessons from the development of ondansetron, aprepitant, ramelteon, varenicline, lorcaserin, and suvorexant. Part I. J Psychiatr Pract. 2014;20:460–465.
    36. Preskorn SH. CNS drug development: lessons learned Part 2. Symptoms, not syndromes as targets consistent with the NIMH research domain approach. J Psychiatr Pract. 2015;21:60–66.
    37. Preskorn SH. Psychiatric and central nervous system drugs developed over the last decade: what are the implications for the field? J Psychiatr Pract. 2017;23:352–360.
    38. Preskorn SH. CNS drug development, lessons learned, part 4: the role of brain circuitry and genes—tasimelteon as an example. J Psychiatr Pract. 2017;23:425–430.
    39. Preskorn S. CNS drug development, lessons learned, part 5: how preclinical and human safety studies inform the approval and subsequent use of a new drug—suvorexant as an example. J Psychiatr Pract. 2018;24:104–110.

    locus coeruleus; central adrenergic system; alpha-2 adrenergic agonist; dexmedetomidine; norepinephrine

    Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.