Authors: Mary Beth Bauer, Kevin P.M. Currie
Categories: Article, Adrenal medulla, adrenal cortex, catecholamine, chromaffin, G protein coupled receptor, serotonin, serotonin transporter, serotonylation, sympathetic nervous system, stress
Source: Vitamins and hormones
Authors: Mary Beth Bauer, Kevin P.M. Currie
The adrenal glands are key components of the mammalian endocrine system, helping maintain physiological homeostasis and the coordinated response to stress. Each adrenal gland has two morphologically and functionally distinct regions, the outer cortex and inner medulla. The cortex is organized into three concentric zones which secrete steroid hormones, including aldosterone and cortisol. Neural crest-derived chromaffin cells in the medulla are innervated by preganglionic sympathetic neurons and secrete catecholamines (epinephrine, norepinephrine) and neuropeptides into the bloodstream, thereby functioning as the neuroendocrine arm of the sympathetic nervous system. In this article we review serotonin (5-HT) and the serotonin transporter (SERT; SLC6A4) in the adrenal gland. In the adrenal cortex, 5-HT, primarily sourced from resident mast cells acts as a paracrine signal to stimulate aldosterone and cortisol secretion through 5-HT4 / 5-HT7 receptors. Medullary chromaffin cells contain a small amount of 5-HT due to SERT-mediated uptake and express 5-HT1A receptors which inhibit secretion. The atypical mechanism of the 5-HT1A receptors and interaction with SERT fine tune this autocrine pathway to control stress-evoked catecholamine secretion. Receptor-independent signaling by SERT /intracellular 5-HT modulates the amount and kinetics of transmitter release from single vesicle fusion events. SERT might also influence stress-evoked upregulation of tyrosine hydroxylase transcription. Transient signaling via 5-HT3 receptors during embryonic development can limit the number of chromaffin cells found in the mature adrenal gland. Together, this emerging evidence suggests that the adrenal medulla is a peripheral hub for serotonergic control of the sympathoadrenal stress response.
In mammals, the adrenal glands sit atop the kidneys and act as a hub in the endocrine system, key to maintaining homeostasis and responding to acute stressors. Weighing approximately 4–5 gm each in adult humans, they play a critical role in blood pressure, immune response, generation of sex hormone precursors as well as the flight or fight response. The adrenal glands are comprised of two embryologically distinct tissues, the cortex and the medulla, and an outermost capsule of mesenchymal cells (figure 1). The arrangement of a tightly clustered adrenal medulla surrounded by the adrenal cortex is unique to mammals (Pohorecky & Wurtman, 1971). In addition to providing structure, the capsular layer, which includes fibroblast-like cells and progenitor cells, acts as a source of signals for adrenal development, homeostasis, and regeneration (Kim & Choi, 2020; Lyraki & Schedl, 2021; Pignatti & Flück, 2021; Vidal et al., 2016).
The cortex, which accounts for 80–90% of the adrenal gland, is of mesodermal origin and in humans contains three concentric, histologically distinct layers or functional the outer zona glomerulosa, the middle zona fasciculata and the inner zona reticularis. The zona glomerulosa accounts for approximately 15% of the cortex and is responsible for the production of mineralocorticoids, predominantly aldosterone, which regulates blood pressure and fluid volume by stimulating reabsorption of sodium and thus water. Cells in the zona glomerulosa are arranged in round glomeruli which are surrounded by, and connected through, a basement membrane (Leng et al., 2021). The glomeruli each contain rosettes, a formation of at least five epithelial cells that converge to a single center. Zona glomerulosa cells produce aldosterone primarily in response to angiotensin II and high potassium levels. Aldosterone secretion may also be elicited by adrenocorticotrophic hormone (ACTH), and can be affected by dopamine, atrial natriuretic peptide, serotonin and other neuromediators (Spät & Hunyady, 2004).
Adjacent to the zona glomerulosa, lies the zona fasciculata, which is the largest portion of the adrenal cortex (approximately 75%). Cells of the zona fasciculata are arranged in centripetally oriented radial cords (fascicula) which are surrounded by fenestrated capillaries. ACTH released from the pituitary stimulates production of glucocorticoids by zona fasciculata cells, mainly cortisol in humans and corticosterone in rodents, which exert diverse effects including stimulation of gluconeogenesis, suppression of the immune response, and increasing blood pressure.
The innermost layer of the adrenal cortex, the zona reticularis, appears years after birth in humans; islands of zona reticularis cells become morphologically discernable around three years of age, form a continuous and functional layer around 6 years, and further develop until age 12–13 (Dhom, 1973; Dumontet & Martinez, 2021). Cells of the zona reticularis are organized into a net-like structure and produce androgens, most notably dehydroepiandrosterone (DHEA) and DHEA-sulfate (DHEA-S).
The adrenal medulla, approximately 10–20% of the adrenal gland, is of neuroectodermal origin and chromaffin cells constitute the vast majority of the cells, along with some ganglionic cells, glial-like sustentacular cells, and immune system cells such as resident macrophages (González-Hernández et al., 1994). During embryonic development, neural crest cells migrate to form a sympathoadrenal precursor population near the dorsal aorta, which then further migrate either into the developing adrenal cortex to populate the inner medulla or to form sympathetic ganglia. Therefore, the neural crest-derived medullary chromaffin cells are closely related to post-ganglionic sympathetic neurons (Huber et al., 2009). In addition to this classical understanding of chromaffin cell development, recent evidence shows that neural crest-derived Schwann cell precursors (SCPs) migrate along the splanchnic nerve fibers into the developing adrenal medulla (Furlan et al., 2017). Some of these multipotent SCPs generate a transient population of “bridge” cells (around embryonic days E11.5 – E14.5 in mice) which are the immediate precursors of chromaffin cells. This second wave of migration is estimated to account for over half of the chromaffin cells in the mature adrenal medulla (Bechmann et al., 2021; Kastriti et al., 2020).
Chromaffin cells are arranged in small columns next to numerous sinusoids lined by fenestrated epithelium and secrete catecholamines, epinephrine and norepinephrine, as well as co-transmitters into the bloodstream. They are predominantly adrenergic, i.e., secrete epinephrine (adrenaline), as these cells express phenylethanolamine-N-methyltransferase (PNMT) which converts norepinephrine into epinephrine. Noradrenergic chromaffin cells lack this enzyme and thus secrete norepinephrine. Chromaffin cells also contain a cocktail of other neuropeptides and co-transmitters including neuropeptide Y (NPY), atrial natriuretic factor, vasopressin, Substance P, endogenous opioids (enkephalins, endorphins), and ATP (Hook et al., 2008; Winkler et al., 1986). Of particular relevance to this chapter, chromaffin cells also contain small amounts of serotonin, discussed in more detail in section 4 (figure 1).
The arrangement of an inner medulla surrounded by the outer cortex seen in mammals has implications for crosstalk between the cortical and chromaffin cells. Indeed, numerous bidirectional paracrine interactions between the cortex and the medulla have been identified as crucial to the gland’s development, maintenance and response to stress (Bornstein et al., 1997; Eiden & Jiang, 2018). This communication might be facilitated by some degree of intermingling as chromaffin cells singly, in clusters, or rays, have been observed in the three zones of the cortex of adult rat (Bornstein et al., 1991; Bornstein et al., 1994; Gallo-Payet et al., 1987; Palacios & Lafarga, 1975). Interestingly, cortical cells have been observed singly, in clusters or islets in the adrenal medulla (Bornstein et al.; Bornstein et al., 1994). Crosstalk between the different adrenal zones might also be facilitated by the extensive vascularization within the gland.
The adrenal gland is extremely vascularized and compared to other tissues has one of the highest basal blood flow rates in proportion to its size (Williams & Leggett, 1989). Three main arteries supply the adrenal gland with the superior adrenal artery which is a branch of the inferior phrenic artery, middle adrenal artery which branches directly from abdominal aorta, and inferior adrenal artery which branches from the renal artery bilaterally. The blood vesselss branch to create a plexus in and under the capsule, from which the capsule, cortex and medulla are supplied. Cortical arterioles, the arteriae cortices, descend to form a network in the zona glomerulosa and fenestrated sinusoids through the zona fasciculata; these vessels merge near the cortico-medullary boundary as a capillary network to carry blood to medullary capillary fenestrated sinusoids. Other branches of the capsular plexus, the arteriae medullae, pass directly through the cortex to deliver blood to medullary capillary sinusoids. Thus, the adrenal medulla is perfused both by medullary arterioles providing oxygen- and nutrient- rich blood and by cortical capillaries delivering high concentrations of steroid hormones. Venules from the adrenal sinusoids converge on a central medullary vein in each adrenal gland; the right central adrenomedullary vein drains directly into the inferior vena cava, and the left central adrenomedullary vein drains directly into the left renal vein.
The adrenal gland and its capsule are predominantly innervated by fibers from the greater splanchnic nerve and to a lesser extent, the paravertebral and suprarenal ganglia of the sympathetic nervous system and the vagus nerve. Most of the fibers of the splanchnic nerve are preganglionic sympathetic neurons, with a much smaller portion being postganglionic. Splanchnic nerve bundles penetrate the capsule and form a subcapsular network from which they cross through the cortex, mostly in association with blood vessels and connective tissue trabeculae, to synapse in the medulla.
While most splanchnic nerve fibers synapse in the medulla, some appear to terminate in the cortex and along blood vessels. Cholinergic fibers with varicosities in the cortex have been observed, with a presumed origin from the preganglionic sympathetic splanchnic supply (Charlton et al., 1991; Jánossy et al., 1998), although a postganglionic parasympathetic innervation has also been hypothesized (Jánossy et al., 1998). Postganglionic sympathetic fibers have been associated with cells in the superficial cortex(Tóth et al., 1997). These catecholaminergic fiber varicosities make close contact with zona glomerulosa cells and blood vessels (Charlton et al., 1992; Holzwarth et al., 1987; Tóth et al., 1997; Vizi et al., 1992), and are believed to modulate blood flow (Carlsson et al.).
Various neuropeptides, transmitters and enzymes have been associated with cortical fibers, often in association with the subcapsular plexus, blood vessels and zona glomerulosa cells. Some of these might reflect contributions from adrenal medullary ganglion cells which have been hypothesized to provide intrinsic, intra-adrenal innervation(Díaz-Flores et al., 2008; Nussdorfer, 1996; Parker et al., 1993; Unsicker et al., 1978); indeed, ganglion cells have been described as containing one or more of the following transmitters or noradrenaline, NPY, VIP, nitric oxide synthase, acetylcholine, and GABA (Afework et al., 1995; Holgert et al., 1995; Holgert et al., 1996; Kato et al., 2014; Kondo, 1985; Oomori et al., 1994).
The adrenal medulla, which comprises the neuroendocrine arm of the sympathetic nervous system, is predominantly innervated by sympathetic preganglionic neurons. These neurons originate from the intermediolateral cell column of the thoracic spinal cord and travel to the adrenal gland via the splanchnic nerve (Kesse et al. 1988). After penetrating the capsule and traversing the adrenal cortex, preganglionic fibers synapse on adrenal chromaffin cells in the medulla. Accordingly, the adrenal medulla is analogous to a sympathetic ganglion and the chromaffin cells to modified postganglionic sympathetic neurons. The preganglionic neurons release acetylcholine (ACh) and neuropeptides (e.g. pituitary adenylyl cyclase activating peptide [PACAP]) to stimulate chromaffin cells (Carbone et al., 2019; Smith & Eiden, 2012; Stroth et al., 2013). In turn, the chromaffin cells secrete catecholamines, primarily epinephrine and norepinephrine, as well as neuropeptides and co-transmitters into the circulation. The circulating catecholamines drive numerous responses across the cardiovascular, endocrine, immune, and nervous systems which help maintain physiological homeostasis and evoke the “fight or flight” response to acute stress. This sympathetic neuroendocrine response complements the postganglionic sympathetic neurons which provide noradrenergic innervation to specific target organs.
The firing rate of sympathetic preganglionic neurons is low at rest (<1Hz) and controlled by descending input from the brainstem, primarily from the rostral ventrolateral medulla (RVLM) along with the rostral ventromedial medulla, A5 noradrenergic cell group, paraventricular nucleus of the hypothalamus, and the caudal raphe nuclei (Bacon et al., 1990; Jansen et al., 1995; Kerman et al., 2006; Loewy, 1981; Strack et al., 1989; Verberne et al., 2014) (also see section 4 below). These brain regions integrate various sensory, emotional, and other inputs to coordinate central drive to the peripheral sympathetic nervous system, which can increase substantially (> 10–15Hz) during periods of intense stress. Notably, innervation of the adrenal gland by the splanchnic nerve is not fully developed at birth. In spite of this, neonatal chromaffin cells respond to hypoxia, presumably to provide catecholamines critical to adapting to extra-uterine life (Nurse et al., 2018). This intrinsic sensitivity to hypoxia decreases as the splanchnic innervation of the adrenal medulla matures postnatally.
In addition to the predominant innervation by preganglionic sympathetic neurons, the medulla also receives postganglionic sympathetic innervation (Kesse et al., 1988; Parker et al., 1990). Limited data suggest a sparse innervation from the dorsal motor nucleus of the vagus (Coupland et al., 1989; Dun et al., 1996; Parker et al.). Afferent nerve fibers projecting from the adrenal medulla to sensory ganglia have also been reported; these medullary sensory nerve fibers project to ipsilateral dorsal root ganglia, primarily at thoracic levels (Dun et al., 1996; Kumar et al., 2010; Mohamed et al., 1988; Parker et al., 1993; Qi et al., 1991) and are capsaicin-sensitive (Dun et al., 1996; Ulrich-Lai et al., 2002). Other sensory vagal afferents project to the ipsilateral and contralateral nodose ganglia (Coupland et al., 1989; Dun et al., 1996; Parker et al., 1993).
Serotonin, or 5-hydroxytryptamine (5-HT), is an evolutionarily ancient monoamine transmitter which modulates a diverse array of physiological processes in the central nervous system and peripheral tissues of mammals. 5-HT is synthesized from the essential amino acid tryptophan in a two-step process; tryptophan hydroxylase (TPH) catalyzes the rate-limiting step converting tryptophan to 5-hydroxytrypotophan (5-HTP), followed by aromatic amino acid decarboxylase (AADC) which converts 5-HTP into 5-hydroxytryptamine (5-HT) (figure 2). There are two isoforms of tryptophan hydroxylase which display distinct expression patterns and kinetics. TPH1 is primarily expressed in the enterochromaffin cells lining the gastrointestinal tract, the pineal gland, and a few other peripheral tissues including lungs, pancreas, and fat (Matthes & Bader, 2018). The enterochromaffin cells account for production of the vast majority of 5-HT in the body (>90%). Platelets in the blood contain a significant pool of peripheral 5-HT, but rather than being synthesized, this supply is due to uptake by the serotonin transporter (SERT) as the platelets transit through the gut. The platelet 5-HT is then stored in dense core granules for use in coagulation, hemostasis, and distribution throughout the body. There is little passage of 5-HT across the blood-brain barrier, so the “central” and “peripheral” serotonin systems are independent from one another. Underscoring this, central serotonergic neurons clustered in the raphe nuclei of the brainstem primarily use TPH2 (rather than TPH1) to synthesize 5-HT, as do serotonergic neurons of the enteric nervous system. Although there are a relatively small number of serotonergic neurons in the raphe nuclei, they project widely throughout the brain and spinal cord (Bang et al., 2012; Hale et al., 2012; Kerman et al., 2006; Kinney et al., 2009; Moore et al., 1978) enabling complex serotonergic modulation of many cognitive and behavioral functions including mood, social interaction, anxiety, sleep, appetite, and autonomic regulation.
Once synthesized, 5-HT is pumped into synaptic or dense core vesicles by the vesicular monoamine transporter (VMAT1/2) where it is stored for subsequent use as a transmitter. Alternatively, 5-HT is degraded by the sequential actions of monoamine oxidase (MAO) and aldehyde dehydrogenase to 5-hydroxyindoleacetic acid (5-HIAA) for excretion in the urine. Another key mediator of 5-HT signaling and homeostasis is the serotonin transporter (SERT; SLC6A4). SERT-mediated reuptake is one of the main mechanisms for clearing 5-HT from the synaptic cleft, thereby terminating receptor activation, and recycling the 5-HT to be packaged into secretory vesicles for future reuse. Genetic variations in SERT have been associated with depression, anxiety, post-traumatic stress disorder (PTSD), obsessive compulsive disorder, and autism (for reviews see (Daws & Gould, 2011; Murphy et al., 2008; Ye & Blakely, 2011)). Indeed, SERT is a clinically relevant target for widely prescribed antidepressants (e.g. selective serotonin reuptake inhibitors), and for psychostimulant drugs (e.g. cocaine and MDMA / ecstasy) (Torres et al., 2003). SERT is also expressed in a variety of peripheral tissues including the gastrointestinal system, platelets, blood vessels, heart, lung, and pancreas (Gershon, 2013; Linder et al., 2009; Martin et al., 2017; Oliver et al., 2016; Penumatsa & Fanburg, 2014; Watts et al., 2012). Chromaffin cells in the adrenal medulla robustly express SERT (Schroeter et al., 1997), and as discussed below, this might coordinate a peripheral hub for control of the sympathoadrenal stress response (Brindley et al., 2017).
5-HT can exert its effects on target cells through a large (14 member) and diverse family of receptors which are classified into seven major sub-families (5-HT1–7) (Barnes et al., 2021). Six of the seven 5-HT receptor families are G protein coupled receptors (GPCRs), with the exception being 5-HT3 receptors which are cys-loop ligand-gated ion channels (similar to nicotinic acetylcholine receptors). The GPCR family members couple to all the major G-protein dependent signaling pathways; the 5-HT1 and 5-HT5 families couple primarily to Gi/o-type G proteins, the 5-HT2 family couples to Gq, and the 5-HT4, 5-HT6, and 5-HT7 families couple to Gs. The intracellular signaling repertoire of the receptors extends beyond canonical G protein signaling cascades and includes signaling through Gβγ-dimer, GRKs and β-arrestin pathways. 5-HT can also signal via a novel receptor-independent mechanism in which the enzyme transglutaminase-2 can covalently link 5-HT to glutamine residues in target proteins, a process dubbed “serotonylation” (Bader, 2019; Jiang et al., 2021; Muma & Mi, 2015; Walther et al., 2011). The list of target proteins that can be serotonylated is growing and includes small G proteins, actin, fibronectin, Akt, and histone H3. Consequently, serotonylation has been implicated in α-granule release from platelets, modulation of vascular function, insulin granule release from pancreatic beta cells, and transcriptional regulation in neurons (Bader, 2019; Jiang et al., 2021; Walther et al., 2011). Thus, in addition to terminating extracellular signaling pathways, SERT mediated uptake can provide a source of intracellular 5-HT for potential use in these receptor-independent signaling pathways.
Serotonin stimulation of corticosteroid secretion is mediated by 5-HT receptors, requires cAMP production, PKA activation, and calcium influx through T-type (CaV3) voltage-gated calcium channels, with some possible involvement of L-type (CaV1) voltage-gated calcium channels (Contesse et al., 1996; Lenglet et al., 2002; Louiset et al., 2017; Rossier, 2016; Yang et al., 2020).
In frog and humans, 5-HT binds to 5-HT4 receptors (Contesse et al., 1994; Contesse et al., 1996; Idres et al., 1991; Lefebvre et al., 1992; Lefebvre et al., 1993). While in rat, 5-HT7 receptors mediate the serotonergic effects in zona glomerulosa cells (Contesse et al., 1999; Lenglet et al., 2000; Lenglet et al., 2002). In earlier studies, prior to the discovery of 5-HT7 receptor cDNA and protein in the adrenal cortex, ketanserin inhibition of serotonin’s effects prompted consideration of 5-HT2 receptor involvement (Matsuoka et al., 1985; Rocco et al., 1986; Williams et al., 1984).
5-HT7 receptor expression is increased in the adrenal cortex of rats subjected to chronic restraint (Garcia-Iglesias et al., 2013) or chronic corticosterone treatment (Saroj et al., 2019), suggesting it might be involved in the response and/or adaptation of the gland to stress. Elevated expression of eutopic 5-HT4 receptors as well as aberrant expression of 5-HT4, 5-HT6, and 5-HT7 receptors have been observed in adrenocortical tissue from patients with adrenocortical neoplasms or hyperplasia (Bram et al., 2016; Le Mestre et al., 2019; Lefebvre et al., 2015; Mannelli et al., 2003; Ye et al., 2007). Other components of the serotonin-evoked signaling pathway in the adrenal cortex have also been associated with such conditions (Bram et al., 2016; Felizola et al., 2014).
Paracrine sources of serotonin for stimulation of steroid hormone secretion within the adrenal cortex have been proposed. For example, serotonin is well recognized as being present in adrenergic (PNMT expressing) adrenal chromaffin cells in the medulla (see section 4 of this chapter), and possibly, in sparse chromaffin cells present in,or medullary tissue extending into the cortex of rats, mice and frogs (Delarue et al., 1988; Fernandez-Vivero et al., 1993; Holzwarth & Brownfield, 1985; Holzwarth et al., 1984; Verhofstad & Jonsson, 1983). Additionally, perivascular and intravascular mast cells have been cite as a source of serotonin in rat and human adrenal cortex (Bram et al.; Hinson et al., 1989; Lefebvre et al., 2001; Lefebvre et al., 1996; Lefebvre et al., 1992). In vitro, a mast cell degranulator increased serotonin and subsequently corticosteroid secretion (Hinson et al., 1989; Lefebvre et al., 2001; Lefebvre et al., 1996). A few reports suggest the serotonin transporter (SERT) might play a role. SERT-like immunoreactivity was scant in the adrenal cortex of control rats and substantially increased in the cortex of rats subjected to chronic stress; the association of this immunoreactivity in chromogranin A-positive cells suggests they were likely (medullary) chromaffin cells present in the cortex (Shanker et al., 2020). Similarly, 5-HT-like immunoreactivity observed as clusters in the adrenal cortex was increased in rats subjected to chronic restraint stress compared to sparse immunoreactivity in controls (Garcia-Iglesias et al., 2013).
The presence of tryptophan hydroxylase (TPH), the rate limiting enzyme for serotonin synthesis, in the adrenal cortex is not well characterized. No TPH protein was observed in rats (Garcia-Iglesias et al., 2013) and no TPH immunostaining was observed in human adrenal cortex (Meyer & Brinck, 1999). However, some reports indicate it might be increased in pathophysiological conditions. For example, TPH was overexpressed in the cortex of human adrenal tissue from patients with high ACTH levels (due to 21-hydroxylase enzyme deficiency or ACTH-dependent Cushing syndrome)(Le Mestre et al., 2019). Also, TPH was aberrantly expressed in human adrenal tissue from patients with primary pigmented nodular adrenocortical disease(Bram et al., 2016). Metabolism of serotonin may occur as monoamine oxidase (MAO)-A and MAO-B are present in the adrenal cortex of humans, rats, and cow, though the relative proportions and zonation of these enzymes varies between species(Carmichael & Pfeiffer, 1985; King et al., 1999; Lenzen et al., 1987; Ramonet et al., 2003; Rodríguez et al., 2000). However, MAO-B has not always been observed in the human cortex, and evidence suggests that MAO-A-immunoreactive cells in the cortex are medullary/chromaffin cells (Lefebvre et al., 2001; Lefebvre et al., 1996).
Taken together, these observations suggest that under normal physiological conditions, serotonin in the adrenal cortex is primarily derived from resident mast cells and possibly medullary chromaffin cells, and acts in a paracrine manner to enhance secretion of aldosterone and cortisol/corticosterone through 5-HT4 (human) and 5-HT7 receptors (rat). Serotonin might be subject to reuptake and metabolism by SERT and MAO respectively, although these proteins seem to be primarily expressed in the medullary chromaffin cells. Reports indicating upregulation of 5-HT receptors, TPH, and perhaps SERT, raise the possibility that serotonergic signaling is altered and might be involved in the adaptations or plasticity occurring within the adrenal cortex during chronic stress and/or pathophysiological states.
Alterations in autonomic nervous system function, including the sympathoadrenal stress response, have been linked to diseases with disrupted serotonergic signaling including depression and anxiety (Bedi & Arora, 2007; Caspi et al., 2003; Coughlin, 2011; Gold et al., 2005; Paine et al., 2015; Pervanidou & Chrousos, 2010; Southwick et al., 1999; Weinstein et al., 2010). Experimental evidence in humans and animal models also implicates SERT / 5-HT in the sympathetic stress response. For example, the counterregulatory response to hypoglycemia includes robust stimulation of epinephrine secretion from adrenal chromaffin cells, and this is enhanced by selective serotonin reuptake inhibitor (SSRI) treatment in humans (Briscoe, Ertl, Tate, & Davis, 2008; Briscoe, Ertl, Tate, Dawling, et al., 2008) and rodents (Sanders et al., 2008). Similarly, the increase in plasma epinephrine concentration produced by restraint stress in rodents is enhanced in SERT knockout mice (SERT^−/−^ mice) (Murphy & Lesch, 2008; Tjurmina et al., 2002; Tjurmina et al., 2004). In all of these cases, SERT function throughout the entire body is compromised (i.e. gene knockout or pharmacological block), leaving the cellular targets and mechanism(s) of action unclear. These might include SERT / 5-HT signaling in the brain and/or spinal cord to modulate central drive to the peripheral sympathetic nervous system, along with more direct effects due to SERT / 5-HT signaling within the adrenal medulla (Brindley et al., 2017) (figure 3).
The preganglionic sympathetic neurons that innervate adrenal chromaffin cells are located in the intermediolateral cell column of the thoracic spinal cord (Kesse et al., 1988). Basal firing rates are typically low (<1Hz) but rise several fold (as high as 15–30Hz) during periods of intense stress. Excitatory drive to the preganglionic neurons is primarily from the rostral ventrolateral medulla (RVLM) in the brainstem with direct input also coming from the rostral ventromedial medulla, A5 noradrenergic cell group, paraventricular nucleus of the hypothalamus, and the serotonergic caudal raphe nuclei (Bacon et al., 1990; Jansen et al., 1995; Kerman et al., 2006; Loewy, 1981; Strack et al., 1989; Verberne et al., 2014) (figure 3). Serotonergic neurons are clustered in the raphe nuclei of the brainstem and project extensively throughout the brain and spinal cord (Bang et al., 2012; Hale et al., 2012; Kerman et al., 2006; Kinney et al., 2009; Moore et al., 1978) (figure 3). 5-HT can modulate firing of sympathetic premotor neurons in the RVLM, being both inhibitory (likely via 5-HT1A receptors) and excitatory (likely via 5-HT2 receptors) (Bago et al., 1999; Helke et al., 1997; Horiuchi et al., 2008; Miyawaki et al., 2001; Ramage & Villalon, 2008). In the spinal cord, the preganglionic sympathetic neurons are innervated by SERT-expressing nerve terminals (Appel et al., 1986; Bacon & Smith, 1988; Holets & Elde, 1982; Sur et al., 1996) likely reflecting projections from the caudal raphe nuclei and a small population of spinal serotonergic neurons (Llewellyn-Smith et al., 2006; Newton & Hamill, 1989) (figure 3). 5-HT can directly excite the preganglionic sympathetic neurons via 5-HT2 receptors, or produce disinhibition of spinal interneurons (Lewis & Coote, 1990; Lewis et al., 1993; Madden & Morrison, 2006, 2008; Pickering et al., 1994; Zimmerman et al., 2012). Thus SERT / 5-HT can modulate sympathetic drive to the adrenal medulla at multiple sites within the brain and spinal cord.
Recently, we proposed that chromaffin cells in the adrenal medulla might be a previously unrecognized hub at which peripheral 5-HT controls the sympathoadrenal stress response (Brindley et al., 2017). SERT is robustly expressed in the adrenal medulla, specifically in epinephrine containing chromaffin cells (Schroeter et al., 1997). However, in contrast its well-known roles in CNS serotonergic neurons, SERT / 5-HT signaling in adrenal chromaffin cells, which synthesize and secrete catecholamines, is only beginning to be unraveled.
Mammalian adrenal chromaffin cells do not synthesize 5-HT as they lack tryptophan hydroxylase, the rate-limiting biosynthetic enzyme. However, they do accumulate some 5-HT due to SERT-mediated uptake (Brindley et al., 2016; Holzwarth et al., 1984; Kent & Coupland, 1984; Linder et al., 2009; Meyer & Brinck, 1999; Verhofstad & Jonsson, 1983). The bloodstream is a likely source of 5-HT, as the gland is highly vascularized and blood contains a significant amount of 5-HT in platelets (Mercado & Kilic, 2010; Watts et al., 2012). Indeed, when 5-HT is infused into rodent circulation there is prominent SERT-mediated uptake in the adrenal medulla (Linder et al., 2009). Other potential sources of 5-HT include resident mast cells in the adrenal cortex (Lefebvre et al., 2001), or possibly sensory and vagal nerve terminals in the medulla, some of which might contain 5-HT (Fernandez-Vivero et al., 1993). Consistent with an important role for SERT, the amount of 5-HT in the adrenal glands of SERT^−/−^ mice or rats is reduced by ≈ 65 – 80 % (Brindley et al., 2016; Linder et al., 2009; Tjurmina et al., 2002). One caveat to interpreting these experiments is that body-wide knockout of SERT depletes 5-HT from platelets resulting in reduced delivery via the circulation. Recently, a Cre-loxP approach was used to develop a tissue-specific knockout of SERT from the sympathoadrenal system (SERT^ΔTH^ mice) (Brindley et al., 2019). Adrenal 5-HT content was reduced by ≈ 50 % in SERT^ΔTH^ mice even though blood 5-HT was unaltered, confirming that SERT expressed in adrenal chromaffin cells is an important determinant of adrenal 5-HT content and homeostasis. The 5-HT remaining in the adrenal glands from SERT^ΔTH^ mice might reflect residual blood in the isolated glands and mast cells in the adrenal cortex. There is also evidence for SERT-independent uptake of 5-HT by the adrenal medulla (Linder et al., 2009). The underlying transporter(s) remains unclear but might include members of the organic cation transporter family, which are known to be expressed in rodent and human adrenal gland tissue (Béery et al., 2003; Couroussé et al., 2015; Duan & Wang, 2010), or the plasma membrane monoamine transporter (PMAT; SLC29A4) which might be important for accumulating 5-HT used during development of adrenal chromaffin cells (Kameneva et al., 2022).
Following SERT-mediated uptake into adrenal chromaffin cells, some of the intracellular 5-HT is transported into secretory vesicles by vesicular monoamine transporters (VMAT1/2) and thereby protected from being metabolized in the cytosol by monoamine oxidase (figure 4A). Vesicular catecholamine concentrations approach 0.5–1 molar (Albillos et al., 1997; Winkler & Westhead, 1980), so even though 5-HT is much less abundant than epinephrine (< 0.15%) (Brindley et al., 2016), there is still a significant amount present and presumably available for exocytosis. Therefore, 5-HT might act in an autocrine or paracrine manner within the adrenal medulla to control chromaffin cell function (figure 4A). This was investigated using carbon fiber amperometry, an extremely sensitive electrochemical technique that can resolve transmitter release from individual vesicles that fuse with the plasma membrane (Borges et al., 2008; Jewell & Currie, 2013; Mosharov & Sulzer, 2005; Wightman et al., 1991). The catecholamine released by a vesicular fusion event is oxidized by the carbon fiber electrode resulting in a transient “spike” on the amperometric current recording. The number of spikes reflects how many vesicles undergo exocytosis, while other spike properties provide detailed insight into the amount and kinetics of catecholamine secretion from these unitary events. These experiments revealed that 5-HT can inhibit catecholamine secretion from isolated adrenal chromaffin cells and that this is most likely mediated via 5-HT1A receptors; it is blocked by a 5-HT1A receptor antagonist and mimicked by a selective 5-HT1A receptor agonist, 8-OH-DPAT (Brindley et al., 2016) (figure 4A).
The 5-HT1A receptors couple to Gi/o-type G proteins and the resultant G protein βγ (Gβγ) subunit signaling reduced the number of secretory vesicles that underwent exocytosis. Analyzing the individual spike parameters showed that 5-HT1A receptors did not alter the amount or kinetics of transmitter release from individual fusion events. The precise downstream mechanism is distinct from other inhibitory GPCRs on chromaffin cells. ATP, endogenous opioids, and catecholamines produce autocrine inhibition via P2Y purinergic receptors, μ opioid receptors, and α2 adrenergic receptors, respectively (Albillos et al., 1996; Brede et al., 2003; Carabelli et al., 1998; Currie & Fox, 1996). Prostaglandin EP3 receptors also inhibit secretion, perhaps allowing for paracrine signaling from resident immune system cells in the adrenal (Currie et al., 2000; Jewell et al., 2011). The main mechanism of all these Gi/o-coupled receptors is through inhibition of voltage-gated calcium channels by the Gβγ subunits which decreases Ca^2+^ entry and thus vesicle exocytosis (Bauer & Currie, 2020; Carbone et al., 2019; Jewell & Currie, 2013; Zamponi & Currie, 2013). In the brain, 5-HT1A receptors have been reported to reduce cellular excitability and neurotransmitter release by activating K^+^ channels and/or inhibiting voltage-gated Ca^2+^channels (Bayliss et al., 1997; Courtney & Ford, 2016; Foehring, 1996; Luscher et al., 1997; Penington & Kelly, 1990; Penington et al., 1993). However, in chromaffin cells the inhibition of secretion by 5-HT1A receptors does not involve Ca^2+^ channels, K^+^ channels, or changes in intracellular [Ca^2+^] (Brindley et al., 2016). Other potential targets for Gβγ-mediated inhibition of transmitter release might include SNARE proteins, which comprise the core exocytotic fusion machinery, or dynamin (Chen et al., 2005; Gerachshenko et al., 2005; Yoon et al., 2008; Zurawski et al., 2016; Zurawski et al., 2019).
We have previously proposed that the unique features of serotonergic signaling make it well-suited to control catecholamine secretion during periods of intense stimulation or stress (Brindley et al., 2017) (figure 5). The inhibition of Ca2+ channels by P2Y and other GPCRs can be reversed by strong depolarization / repetitive firing (Currie & Fox, 2002; Zamponi & Currie, 2013). This reversal, along with build-up of residual Ca2+ within the cells during intense / sustained stimulation, is expected to diminish the effectiveness of autocrine inhibition by GPCRs that target Ca2+ channels. In contrast, the distinct Ca2+ independent mechanism of 5-HT1A receptors will persist under these conditions. SERT function adds a further layer of complexity by limiting the ability of 5-HT to recruit the inhibitory signaling pathway (Brindley et al., 2016; Brindley et al., 2017). Presumably, SERT-mediated uptake reduces the availability of extracellular 5-HT to activate the 5-HT1A receptors. Consistent with that, inhibition by lower concentrations of 5-HT is only apparent in cells isolated from wild-type mice if SERT function is pharmacologically blocked (with escitalopram), or in cells isolated from SERT^−/−^ mice (Brindley et al., 2016). It is interesting to speculate that this “gatekeeping” role of SERT might prevent activation of 5-HT1A receptors under basal stimulation, but that during periods of intense stimulation the amount of extracellular 5-HT increases enough to overwhelm SERT and activate the 5-HT1A receptors (Brindley et al., 2017)(figure 5).
Some aspects of the model outlined above for autocrine control of adrenal chromaffin cells by SERT / 5-HT1A receptors still need to be verified experimentally. However, it is clear that SERT plays a pivotal role. The 5-HT content of adrenal chromaffin cells is due to SERT-mediated uptake and the transporter limits activation of 5-HT receptors, at least by low-moderate concentration of 5-HT. Consequently, loss of SERT function might have complex effects depending on timing and context. Acute pharmacological block of SERT might enhance serotonergic inhibition whereas chronic loss or block of SERT might have the opposite effect. In SERT^−/−^ mice, the 5-HT content of chromaffin cells is dramatically reduced, which is predicted to preclude autocrine serotonergic inhibition. This loss of local control within the adrenal medulla could contribute to the enhanced plasma epinephrine concentration produced by restraint stress in SERT^−/−^ mice (Brindley et al., 2017; Murphy & Lesch, 2008; Tjurmina et al., 2002; Tjurmina et al., 2004) or by hypoglycemia in rats and humans treated with SSRIs (Briscoe, Ertl, Tate, & Davis, 2008; Briscoe, Ertl, Tate, Dawling, et al., 2008; Sanders et al., 2008).
In addition to the inhibitory effects of 5-HT1A receptors outlined above, carbon fiber amperometry experiments revealed another effect of SERT / 5-HT signaling on catecholamine secretion from chromaffin cells (Brindley et al., 2016). In the absence of receptor activation, there was no difference in the number or timing of amperometric spikes evoked by membrane depolarization when comparing wild-type and SERT^−/−^ chromaffin cells. However, the charge of the spikes (i.e. quantal size) was significantly smaller (35%) and spike duration was significantly shorter in SERT^−/−^ cells (figure 4B). A recent preliminary report confirms this observation; quantal size was smaller in chromaffin cells isolated from SERT knockout and this was mimicked in wild type cells by treating with escitalopram (an SSRI antidepressant which blocks SERT) for > 24 hours, confirming it was not due to compensatory changes in the knockout mice (Currie et al., 2022). Therefore, it appears that SERT / 5-HT signaling has two mechanistically distinct effects on secretion from adrenal chromaffin cells. First, 5-HT1A~ receptors inhibit secretion by reducing the number of vesicles that fuse with the plasma membrane (figure 4A). Second, loss of SERT function (genetic deletion or pharmacological block in wild-type cells) leads to reduced quantal size (spike charge) and faster release kinetics (spike duration) with no effect on the number of fusion events (figure 4B).
How SERT function controls quantal size remains unclear. One possibility is that the amount of transmitter packaged per individual vesicle is reduced in SERT−/− mice. However, the catecholamine content of adrenal glands from SERT−/− mice is no different than wild-type and even though 5-HT is dramatically reduced, this only accounts for a very small fraction (<0.15 %) of the total monoamines. Another possibility is that SERT / 5-HT can modulate the fraction of the vesicle cargo that is released during exocytosis (figure 4B). Rather than being an all-or-none event, it is now apparent that transmitter release from large dense core vesicles in chromaffin cells (and perhaps synaptic vesicles) is much more complex and can result in complete or only partial release of the vesicle cargo. This might also serve to allow for selective release of smaller transmitters compared to larger peptidergic cargo. One control point for such regulation is the fusion pore that forms between the vesicle and extracellular space to allow for transmitter efflux (Cardenas & Marengo, 2016; Chang et al., 2017). Transient fusion pore opening (for example “kiss-and-run” exocytosis) might favor only partial emptying depending on how long the pore remains open. In chromaffin cells the fusion mode / prevalence of transient fusion events can be influenced by various factors including cellular firing patterns, G proteins, kinases, synaptotagmin isoform, and remodeling of the actin cytoskeleton (Bendahmane et al., 2020; Cardenas & Marengo, 2016; Chen et al., 2005; Elhamdani et al., 2001; Fulop et al., 2005; Fulop & Smith, 2006; Yoon et al., 2008). The vesicle cargo, in particular peptides and the chromogranin “core”, have also been reported to alter the fusion kinetics and quantal size (Mark Wightman et al., 2018; Weiss et al., 2014). This raises the possibility that following uptake by SERT, cytosolic or vesicular 5-HT might favor more complete emptying of vesicle contents, and thus larger quantal size and duration of release events (amperometric spikes). As outlined above, receptor independent signaling by 5-HT due to serotonylation of intracellular proteins has been reported to control a variety of cellular functions, including insulin granule trafficking, actin organization, and neuronal transcription, but whether this novel signaling paradigm contributes to serotonergic control of catecholamine secretion or other chromaffin cell functions remains an open question.
In addition to modulating catecholamine secretion, SERT/5-HT signaling might also help regulate the homeostatic response to acute stress. Tyrosine hydroxylase (TH) is the rate limiting enzyme for catecholamine synthesis, and in response to acute stress its expression and function is increased in the adrenal medulla to replenish and maintain catecholamine stores (Kvetnansky et al., 2013; Nankova et al., 1999; Sabban et al., 1997; Stroth et al., 2013; Wang et al., 2013; Xu et al., 2007). This stress-evoked upregulation of tyrosine hydroxylase is largely dependent on neuronal input and release of not only acetylcholine, but also the neuropeptide PACAP (pituitary adenylyl cyclase activating polypeptide) from preganglionic splanchnic nerve terminals (Mustafa, 2013; Smith & Eiden, 2012; Stroth et al., 2013). In SERT^−/−^ mice, plasma catecholamine levels at rest were similar to wild type but the elevation produced by restraint stress, specifically of epinephrine, was enhanced. In spite of this enhanced secretion of epinephrine, SERT^−/−^ mice fail to upregulate expression of tyrosine hydroxylase, and consequently display diminished adrenal catecholamine content immediately following the acute restraint stress (Armando et al., 2003; Tjurmina et al., 2002; Tjurmina et al., 2004). It is tempting to speculate that local serotonergic signaling (and SERT expression) in the adrenal chromaffin cells is involved in this phenomenon and future experiments using mice with sympathoadrenal specific knockout of SERT (SERT^ΔTH^ mice) (Brindley et al., 2019) will help clarify this. Restraint stress also increases expression of angiotensin II receptors in wild-type mice and this was disrupted in the SERT−/− model (Armando et al., 2003). It will be of interest to determine if SERT function is also involved in upregulation of other neuropeptides in chromaffin cells such as galanin, NPY, and enkephalin.
The mechanism by which SERT might mediate these effects is also unclear. As already mentioned, histone H3 serotonylation is permissive for gene transcription in neurons (Farrelly et al., 2019) but whether it plays a role on chromaffin cells remains unknown. That said, serotonin immunoreactivity has been reported in the nucleus of chromaffin cells (Csaba & Sudár, 1978). This novel mechanism would be an intriguing possibility and given that clinically relevant doses of SSRI antidepressants lead to > 80% occupancy (block) of SERT (Meyer et al., 2004; Owens et al., 2008) it is possible that these drugs could impact the uptake of 5-HT and thus the transcriptional response to acute stress. There is also the question of what role SERT / 5-HT signaling plays in regulating the response to chronic, repetitive stress. While that question remains open, it has been reported that expression of SERT is upregulated in chromaffin cells in the adrenal glands of rats following a chronic restraint stress paradigm (Shanker et al., 2020). This raises the possibility that there is a complex bidirectional relationship between SERT expression and stress at the level of the adrenal gland.
As already outlined, the adrenal cortex and medulla are two embryologically and developmental distinct structures enclosed within a common capsule. The cortex is derived from mesoderm whilst the medulla, including chromaffin cells, are derived from the neural crest. A full discussion of adrenal medulla development is beyond the scope of this article (for recent reviews see (Bechmann et al., 2021; Kastriti et al., 2020), but recent evidence shows there is transient window between mouse embryonic days E11.5 – E14.5 during 5-HT3 receptor signaling helps control the generation of chromaffin cells during organ development (Kameneva et al., 2022).
Migratory neural crest cells give rise to a sympathoadrenal precursor population near the dorsal aorta and it was thought that this subsequently generated all the chromaffin cells of the adrenal medulla along with sympathetic ganglia (Huber et al., 2009). However, recent evidence shows that over half the adrenal chromaffin cells arise from a second, consecutive migration of neural crest derived Schwann cell precursors (SCPs) along the splanchnic nerve fibers into the developing adrenal medulla (Furlan, 2017}. Some of these multipotent SCPs then give rise to a transient population of “bridge” cells which are the immediate precursors of the chromaffin cells. In mice, these “bridge” cells are characterized by expression of 5-HT3 receptors (Htr3a), although the role of these receptors in human “bridge” cells is less clear (Kameneva et al., 2022). Chromaffin cells do not express the 5-HT3 receptor. It has been proposed that 5-HT3 receptor signaling in the “bridge” cells limits the generation of new chromaffin cells during adrenal gland development (Kameneva et al., 2022). The authors provide evidence that existing chromaffin cells within the developing embryonic adrenal gland are a source of 5-HT. As the embryonic chromaffin cells lack tryptophan hydroxylase but express the plasma membrane monoamine transporter, it was suggested that the cells accumulate 5-HT via transporter activity, or perhaps synthesize it from circulating 5-HTP. 5-HT released from the existing chromaffin cells might allow for negative feedback to the precursor “bridge” cells. Indeed, if maternal derived 5-HT was elevated by treating pregnant dams with 5-HTP during the critical embryonic period, the size of the adrenal glands and number of chromaffin cells in the offspring was reduced. This was blocked by a 5-HT3 receptor antagonist and mimicked by a 5-HT3 receptor agonist. Maternal stress, which elevated placental and fetal 5-HT, led to a similar reduction in adrenal medulla size in the progeny. This suggests that serotonergic signaling during a critical window of adrenal gland embryonic development is one potential mechanism by which prenatal stress might influence autonomic / endocrine function of the offspring. Finally, the authors note that while elevated 5-HT reduces chromaffin cell number, the opposite is not true; deficits in 5-HT did not impact chromaffin cell development, consistent with previous reports showing no difference in adrenal catecholamine content in SERT knockout mouse models (Brindley et al., 2016; Brindley et al., 2019).
SERT / 5-HT signaling is emerging as a multifaceted regulator of the sympathoadrenal system. SERT could act in the brain and spinal cord to control central drive to the peripheral sympathetic nervous system, and locally in the adrenal chromaffin cells. Previous work using in vivo pharmacological block or gene knockout of SERT (SERT^−/−^ mice or rats) result in body-wide disruption of SERT function. Recently, we reported development of a novel mouse model with selective excision of SERT from the sympathoadrenal system (SERT^ΔTH^ mice) (Brindley et al., 2019). SERT expression and function remains intact in the CNS of SERT^ΔTH^ mice but was abolished in the adrenal gland. This novel model confirmed that SERT expressed in adrenal chromaffin cells is essential for maintaining wild-type levels of 5-HT which might then be reused in an autocrine manner to modulate catecholamine secretion (Brindley et al., 2019). It will also enable experiments to dissect the role of adrenal SERT expression without confounding loss of expression in other tissues.
SERT function and trafficking to the plasma membrane are regulated by cell signaling and genetic polymorphisms (Baganz & Blakely, 2013; Bermingham & Blakely, 2016; Daws & Gould, 2011; Murphy et al., 2008; Ye & Blakely, 2011). It will be interesting to determine if these factors influence the functional impact of SERT signaling in adrenal chromaffin cells. One potential avenue of research is to leverage transgenic mice engineered to model human variants with altered SERT function (Ye & Blakely, 2011). For example, one mouse model has an autism-associated missense SERT mutation (Gly56Ala) which results in elevated transporter function, hyperserotonemia, deficits in social interactions, repetitive behavior, and altered gastrointestinal function (Margolis et al., 2016; Veenstra-VanderWeele et al., 2012). Investigating if/how such mutations impact adrenal chromaffin cells could provide important insights into the autonomic dysfunction associated diseases in which serotonergic signaling is disrupted.
Receptor-independent signaling by 5-HT is also an emerging field and there is some evidence that intracellular 5-HT signaling might modulate the quantal size of transmitter release from individual secretory vesicles in chromaffin cells. Delineating the mechanism of this novel effect will be important, including whether serotonylation might be involved. 5-HT has been visualized in the nucleus of adrenal chromaffin cells (Csaba & Sudár, 1978), so it is tempting to speculate that this might modulate gene transcription, perhaps through serotonylation of histones as recently reported in neurons (Farrelly et al., 2019).