Contents lists available at SciVerse ScienceDirect
Respiratory Physiology & Neurobiology
jou rn al h om epa ge: www.elsev ier .com/ locate / resphys io l
eview
he paracrine role of 5-HT in the control of gill blood flow�
ernd Pelster ∗, Thorsten Schwertenstitute for Zoology and Center for Molecular Biosciences, University of Innsbruck, Technikerstr.25, Innsbruck, Austria
Storage of serotonin in teleost gill cells has been detected in neurons, polymorphous granular cells andin neuroepithelial cells. Innervation from the glossopharyngeal nerve (first gill arch) and the vagus nerve(all gill arches) carries afferent as well as efferent fibers. This innervation extends to the efferent filamentartery, including the sphincter muscle associated with the efferent filament artery, but except for theAntarctic fish does not reach the afferent filament artery. Serotonergic nerves as well as neuroepithelialcells have been shown to release serotonin, while very little is known about the polymorphous granularcells. The paracrine action of the released serotonin may affect vascular smooth muscle cells and pillarcells, which also contain contractile filaments. Already the earliest functional studies revealed a severeincrease in branchial resistance as a result of serotonin application, combined with an increase in theperfusion of the arterio-venous path and the central sinus spaces of the gills. Pharmacological analysisdemonstrated that this is a serotonin specific effect, which in Antarctic fish is due to activation of the 5-HT2
receptor, while inhibition of the 5-HT1 receptor does not reduce the serotonin induced vasoconstrictionof gill blood vessels. Hypoxic degranulation of serotonergic cells evoked the hypothesis that serotonergicvasoconstriction might result in more even and overall better perfusion of gill lamellae. Microscopicanalysis indicated, however, that perfusion of distal lamellae was reduced after serotonin application.Furthermore, a serotonergic increase in branchial resistance caused a decrease in dorsal arterial oxygensaturation, not an increase as would be expected as a result of a better perfusion of gill lamellae. Adetailed analysis of hypoxic effects on gill perfusion revealed that hypoxia induced changes in gill blood
flow are due to cholinergic effects, but serotonergic influences could not be detected. These observationscontradict the hypothesis that serotonergic vasoconstriction might support hypoxic gas exchange. Thefunctional significance of the serotonergic control of gill blood flow therefore is not yet totally clear.Recent observations indicate that specific inhibitors of serotonin re-uptake accumulate in freshwaterand in estuaries. Considering the negative effect of serotonin on arterial blood oxygenation this maybecome a threat to teleost species.
. Introduction
The teleost gill is a complex, multifunctional structure. Whileas exchange typically comes to mind as the primary function,ills being an interface to the environmental water are equallymportant for ionic regulation, acid-base regulation and nitrogenxcretion. Consequently, gills have attracted attention for a longime and detailed descriptions of gill morphology and ultrastruc-
ure have been presented (Olson, 2002; Wilson and Laurent, 2002),ollowed by extensive studies on ion transport mechanisms, gilllood flow, the control of gill perfusion and the modification of
� This paper is part of a special issue entitled “New Insights into Structure/Functionelationships in Fish Gills”, guest-edited by William K. Milsom and Steven F. Perry.∗ Corresponding author at: Institut für Zoologie, Leopold-Franzens-Universität
gill perfusion under various environmental conditions. To accom-modate the different functions in these tiny organs a sophisticatedvascular structure has been developed, and control of blood flowis achieved by combined neuronal and humoral control systems.Efferent and afferent innervation of gills has been established anddetailed studies revealed cholinergic, adrenergic, purinergic, NOand serotonergic effects on blood flow through the various afferentand efferent blood vessels. Thus, an array of possible mechanisms isavailable to adjust branchial blood flow to changing environmentalconditions. In this study we want to summarize our current knowl-edge on the influence of serotonin (5-hydroxytryptamine, 5-HT) ongill perfusion in teleost fish.
2. Innervation of gills and serotonergic storage sites
Histochemical analysis revealed three specific storage sitesin gill tissue, serotonergic neurons, polymorphous granular cells(PGCs) and neuroepithelial cells (NECs) (Dunel-Erb et al., 1982;
Fig. 1. A: Branchial vasculature in the cod. B: Generalized model of the autonomicnervous system of the branchial vasculature focussing on the sites of serotonin(5-HT) signaling. Arrows in vessels show quantitative blood flow direction andoxygen saturation (black: low; grey: high). AFA, afferent filamentary artery; ALa,afferent lamellar arteriole; AVA, arteriovenous anastomosis; BN, branchial nerve;BV,branchial vein; CATs, circulating catecholamines; CVS, central venous sinus; CNS,central nervous system; EFA, efferent filamentary artery; ELa, efferent lamellar arte-riole; GA, gill arch; NV,nutritional vasculature; PAVA, paralamellar arteriovenousanastomosis; SC, sympathetic chain; SL, (secondary) lamella; Sph, sphincter at thebase of the efferent filamental artery; X, vagus nerve. >< and <> indicate vasocon-strictor and vasodilator effects, respectively (Modified from Nilsson and Sundin,1998; Olson, 2002).
B. Pelster, T. Schwerte / Respiratory Phy
ailly et al., 1989, 1992; Zaccone et al., 1992). While the poly-orphous granular cells did not attract a lot of attention in later
ears so that very little information is available on these cells, NECsave been intensively studied and represent the oxygen sensors ofhe gills, oriented towards the internal circulation, but most likelylso towards the external water. They are extensively discussedeparately in this issue (see contribution of Michael Jonz). A care-ul immunohistochemical analysis revealed that not all NECs areerotonergic, serotonergic and non-serotonergic NECs have beendentified on filaments and lamellae of, for example, zebrafish gills.ECs that did not show 5-HT immune-responsiveness may how-ver be immature NECs, not yet fully differentiated (Jonz and Nurse,003; Saltys et al., 2006).
.1. Serotonergic neurons
Serotonergic neurons are an important source of serotoninithin the gills. The branchial nerve innervating teleost gills is pri-arily formed by branches of the glossopharyngeal nerve (IXth
ranial nerve), extending to the first gill arch, the vagus nerv (Xthranial nerve), extending to all gill arches and fibers originating inhe sympathetic chain (Fig. 1B). It is carrying efferent as well asfferent (sensory) fibers (Nilsson, 1984; Sundin and Nilsson, 2002).erotonergic nerve fibers innervate the proximal part of the effer-nt filamental artery including a sphincter muscle associated withhe efferent filament artery (Fig. 1). This innervation extends to thefferent filamental arterioles and to the central venous sinus systemSundin and Nilsson, 2002). Typically no serotonergic innervationf the afferent filament artery is established.
Using a zebrafish neuron-specific antibody zn-12 and a specificntibody against synaptic vesicle protein 2 (SV2), expressed in neu-ons and endocrine cells, a detailed analysis of the innervation ofebrafish gills was performed (Jonz and Nurse, 2003). Bundles oferve fibers emanating from the branchial nerve of the gill archesnter the gill filaments at their base and protrude to the filament tip.rom these nerve bundles zn-12-IR nerve fibers enter the lamellae.he degeneration of the nerve bundles in explant gill cultures indi-ates the external origin of these nerve fibers, i.e. the cell bodies ofhese neurons are located outside the gills. This external innerva-ion is complemented by intrinsic neurons, i.e. by neurons locatedithin the gills. In the gill filament proximal located multipolareurons extend processes distally along the efferent filament arterynd towards the junction between the efferent filament artery andhe efferent branchial artery. SV2-IR fibers appear to innervate thease of the efferent filament artery, which is the site of (or at leastlose to the site of) the sphincter of the efferent artery. Innerva-ion of this sphincter originates from serotonergic and cholinergiceurons of the proximal filament, but extrinsic innervation by sym-athetic and parasympathetic fibers has also been reported (Baillynd Dunel-Erb, 1986; Dunel-Erb and Bailly, 1986). Nevertheless, thentrinsic innervation of the sphincter appears to play an importantole in the control of sphincter activity (Jonz and Nurse, 2003).
SV2-IR fibers are also located on the efferent branchial arteryf the gill arch. A three-dimensional analysis revealed that someroximal neurons are located more superficially near the efferentlament epithelium, while others are located deeper, beneath thefferent filament artery. These proximal neurons are serotonergic.hile the superficial neurons appear to innervate NECs within the
lament epithelium (Jonz and Nurse, 2003), the deeper ones appearo innervate so-called chain neurons intermingled with the centralenous sinus, continuing to more distal regions (Fig. 2). Staining
roperties of these neurons proved their serotonergic responsive-ess. No difference was detected in the innervation patterns of thelaments and lamellae of the four gill arches of the zebrafish, andven the pseudobranch showed a similar innervation pattern.
A comparative study on trout, zebrafish and goldfish revealedquite similar innervation patterns for zebrafish and goldfish withclearly defined nerve bundles in gill filaments extending into thelamellae, while in trout this pattern was somewhat reduced andnever extended into the lamellae (Saltys et al., 2006).
These recent immunohistochemical observations confirm pre-vious studies by Bailly et al. (Bailly et al., 1989), showingthat antibodies against serotonin and 5-methoxytryptamine stainneurons innervating the proximal part of the efferent arterialvasculature, the filament epithelia, the central venous sinus, andcertain other serotonergic cells of the teleost gill filament. Theresults demonstrate the presence of serotonin containing nerve ter-minals, impinging on vascular smooth muscle in the gills, and thesphincter of the efferent artery (Fig. 1).
342 B. Pelster, T. Schwerte / Respiratory Physiolog
Fig. 2. Intrinsic (left) and extrinsic (right) innervation of the zebrafish gill. Intrinsicneurons are located proximal near the efferent branchial artery (eBA) and are sero-tonergic. Extrinsic innervation originates from the branchial nerve (BN), comprisingtiC
2
lNti2c(
tNta
Fftw
he autonomic innervation of the gills. eFA, efferent filament artery; NEC, neuroep-thelial cells; SPN, superficial proximal nerve cells; DPN, deep proximal nerve cells;hN, chain neurons. (Modified after Jonz and Nurse, 2003, with permission.)
.2. Serotonergic NECs
Serotonergic NECs have been described in close contact to pil-ar cells of trout and goldfish (Coolidge et al., 2008). SerotonergicECs have been identified on the filaments and lamellae of several
eleost species, except for trout, where these cells have not beendentified on the lamellae (Fig. 3; (Zaccone et al., 1996; Saltys et al.,006; Coolidge et al., 2008)). In goldfish, NECs appear to be espe-ially concentrated on the lamellae if exposed to hypoxic conditionsTzaneva and Perry, 2010).
NECs, which are thought to be chemosensory cells (see Jonz,his volume), can be expected to be afferently innervated. While
ECs on filaments appear to be mostly innervated, the innerva-
ion of NECs on the lamellae appears to be species specific (Jonznd Nurse, 2003; Saltys et al., 2006; Zaccone et al., 2006; Coolidge
Trout (P = 23 mm Hg)50
Trairao (P = app rox 13 mm Hg)50
Traira (P = 8.7 mm Hg)50
Goldfish (P = 2.6 mm Hg)50
40
30
20
10
0Fila men t La mella e
5-H
T-IR
cells
(cel
lspe
rmm
filam
ent)
ig. 3. Distribution of 5-HT responsive cells of gill filaments and lamellae of dif-erent teleost species, trout (Oncorhynchus mykiss), goldfish, (Carassius auratus),rairão (Hoplias lacerdae), traira (Hoplias malabaricus) (from Coolidge et al., 2008,ith permission).
y & Neurobiology 184 (2012) 340– 346
et al., 2008). In zebrafish (Danio rerio) NECs have been shownto be innervated, but in traira (Hoplias malabaricus) and trairão(Hoplias lacerdae) for NECs on gill lamellae no innervation couldbe detected (Coolidge et al., 2008). While the afferent innervatedcells are regarded as chemosensory cells, NECs without innervationcould be of paracrine importance. NECs often contain a significantnumber of granules. Degranulation of these cells has been observedduring hypoxia, supporting the hypothesis that these cells servea paracrine function ((Dunel-Erb et al., 1982; Jonz et al., 2004;Coolidge et al., 2008); see below). But it cannot be excluded thatinnervated NECs may also be of paracrine importance. Extrinsic orintrinsic nerve fibers innervating filament NECs in the zebrafishmay be efferent and thus allow for a centrally or locally mediatedneurosecretory role of NECs (Jonz and Nurse, 2003). Neuroepithelialbodies (NEBs), peripheral chemoreceptors of the lung, for example,receive a complex innervation pattern of both afferent and efferentfibers (Adriaensen and Scheuermann, 1993).
5-HT-IR cells have also been located on the gill rakers of borch,goldfish and trout (Sundin et al., 1998; Coolidge et al., 2008). It isspeculated that these cells may be chemosensitive, but a recentstudy suggested that these cells are Merkel-like basal cells of tastebuds (Zachar and Jonz, 2012). Merkel-like cells appear to act asan interneuron in taste chemosensitivity, or in mechanosensation.Serotonin released by these cells in the taste buds may have mod-ulatory effects on taste chemoreceptors.
3. Vasomotor effects of serotonin
Isolated perfused trout gills were shown to extract 80% ofthe initial serotonin concentration from the perfusate (Olson,1998), demonstrating that gills have a high binding capacityfor serotonin and that serotonin appears to be metabolized ingill tissue. Besides the metabolic role of the gills to keep sero-tonin plasma levels low (comparable to the pulmonary uptakemechanisms in many air breathing vertebrates) serotonin wasshown to act on gill perfusion (see Fig. 1). Early studies on theeffect of serotonin on the gill vasculature revealed an increase inbranchial vascular resistance following the application of serotonin(5-hydroxytryptamine) (Östlund and Fänge, 1962; Reite, 1969;Katchen et al., 1976; Fritsche et al., 1992).
Following the extensive and widespread innervation of gillsseveral vascular segments may be under serotonergic control. Anobvious target for the serotonergic response is the sphincter asso-ciated with the efferent filament artery (see Fig. 1). In cod, thesphincter is responsive to ACh infusion (Dunel-Erb and Laurent,1977; Smith, 1977) and to electrical stimulation (Pettersson andNilsson, 1979; Nilsson and Pettersson, 1981), and several morerecent studies demonstrated the cholinergic and serotonergicinnervation of the sphincter as discussed above. In zebrafish themajority of innervation of this area was from intrinsic nerve cells(Jonz and Nurse, 2003). These authors concluded that the innerva-tion by mainly proximal neurons of the gills is indicative for localneuronal control of this muscle. Contraction of the sphincter couldcause a redirection of blood towards the arterio-venous pathwayand/or it could result in a better perfusion of the lamellae.
A second important target would be the pillar cells of the lamel-lae, which provide structural support and allow deoxygenatedblood to pass through the vascular sinus. They contain contrac-tile proteins that allow these cells to contract and increase gillvascular resistance in the central parts of the gill filaments. Con-traction of the pillar cells thus could enhance blood flow through
the respiratory lamellae. Gill denervation has been shown to impairgas transfer and to affect the ventilatory response to changes ingas partial pressures of the water (McKenzie et al., 1991; Burlesonand Smatresk, 2000; McKendry et al., 2001), suggesting that gas
siolog
tn
i(vltascctrbothigoigFisdpotdsavtd
EeattitwadcCtotwcry
alcvea
ir
B. Pelster, T. Schwerte / Respiratory Phy
ransfer in the respiratory lamellae may be influenced by the centralervous system.
Studies focusing on the role of a serotonergic vasoconstrictionn gills do not yet generate a conclusive picture. Sundin et al.Sundin et al., 1995) observed that exogenous 5-HT induced aasoconstriction of the distal efferent artery in trout and opposedamellar recruitment. In 15 out of 18 preparations 1 min after injec-ion of 100 nmol/kg serotonin blood flow stopped in efferent andfferent filament arteries for 0.5–2 min. Preincubation with theerotonergic receptor antagonist methysergide abolished the vaso-onstriction of gill vessels, but it did not abolish the decrease inoeliac arterial blood pressure. This suggests that the effect of sero-onin on blood pressure in the gills was achieved by serotonineceptor activation, while in the coeliac artery it was indirect, possi-ly by an interaction between serotonin and adrenergic pathwaysf cardiovascular control. Cardiac output was not reduced underhese conditions, indicating that blood flow within the gills mustave been redistributed, away from distal portions to more prox-
mal regions. Shunting blood away from the distal lamellae anduiding it to the arterio-venous path with central sinus spacesf the gills would impair gas exchange. In previous studies anncrease in arterio-venous flow has been observed after serotoner-ic stimulation in cod gills (Sundin and Nilsson, 1992). Furthermore,ritsche et al. (Fritsche et al., 1992) reported a significant decreasen dorsal arterial oxygen partial pressure following application oferotonin (50–250 nmol/kg) in trout, and concomitant with thisecrease in arterial PO2 dorsal arterial pressure decreased, whileressure in the ventral aorta increased. Consistent with the studyf Sundin et al. (1995) preincubation with methysergide abolishedhe decrease in dorsal arterial PO2, but did not prevent the drop inorsal arterial blood pressure (Fritsche et al., 1992). These detailedtudies therefore confirm the initial observation that serotoninpplication causes a vasoconstriction in gill blood vessels, but thisasoconstriction did not enhance lamellar perfusion. In contrast,his vasoconstriction reduced blood flow to the lamellae, causing aecrease in dorsal arterial PO2.
Somewhat contradictory results have been obtained for theuropean eel. Intravenous application of serotonin to the Europeanel Anguilla anguilla caused a significant branchio-vasoconstrictionnd thus an increase in branchial resistance, but in contrast to trouthe dorsal arterial PO2 increased (Janvier, 1997). As described forrout serotonin caused a stimulation of ventilatory activity, whichn itself may cause an increase in arterial PO2. This certainly haso be expected for eel, where under resting conditions arterial PO2as extremely low with a value of less than 40 Torr (5.3 kPa). To
ccount for this possibility the experiment was repeated under-tubocurarine to block hyperventilation. But even under theseonditions arterial PO2 increased, although to a lesser extent.onsequently the author concluded that, in contrast to trout, sero-onergic vasoconstriction in eel gills resulted in a better perfusionf the gill lamellae. In contrast to this hypothesis, however, ven-ral aortic pressure was not elevated under these conditions, whichould have to be expected as a result of an efferent branchial artery
onstriction (Janvier, 1997). The difference in the response of arte-ial PO2 to serotonergic stimulation in trout and eel therefore is notet fully explained.
The Antarctic borch Pagothenia borchgrevinki attracted specialttention, because this species is constantly exposed to particu-arly low temperatures, has a low vascular resistance and lacks aompact myocardium. Serotonin was identified as a highly potentasoconstrictor in this species, causing a vasoconstriction in thefferent branchial artery, but also in the afferent branchial artery
nd in the gill arch (Forster et al., 1998).
In an in vivo preparation of the borch, injection of serotoninncreased branchial resistance, reduced PaO2 and cardiac outputemained constant (Sundin et al., 1998). The magnitude of the
y & Neurobiology 184 (2012) 340– 346 343
reduction in PaO2 was dependent on the initial PaO2, the higherthe initial PaO2, the higher reduction achieved by serotonin. In thisstudy two 5-HT receptors have been pharmacologically seperated,5-HT1 and 5-HT2. �-methylserotonin, a 5-HT2 receptor agonist,mimicked the effect of serotonin on gills, and specific inhibi-tion of 5-HT2 receptors (LY53857) blocked all responses, whilespecific inhibition of 5-HT1 receptor had no effect. Interestingly,similar to trout the effect of serotonin on dorsal arterial bloodpressure was not modulated by specific inhibition of either 5-HT1or 5-HT2 receptor. This suggests that in Antarctic borch sero-tonergic constriction of gill vessels is mediated by activation of5-HT2receptor. Noteworthy is the very high sensitivity of borch gillvessels towards serotonin. Compared to trout, where serotonin wasused in concentrations around 100 nmol/kg, in borch a concentra-tion of 0.1 nmol/kg elicited a strong response (Sundin et al., 1995,1998).
As expected 5-HT receptor immunoresponsive cells were foundin the outer walls of efferent filament arteries of all 4 gill archesof borch, but 5-HT-IR cells were also found innervating the outerwalls of afferent branchial arteries. This is in line with the observedconstriction of branchial arteries in borch (Forster et al., 1998).Serotonergic innervation of the afferent filament artery appearsto be a peculiarity of Antarctic notothenioid fishes (Sundin andNilsson, 2002).
Based on the immunohistochemical data serotonin release canalso be expected to occur in proximity of the pillar cells. Pillarcells contain contractile proteins, and pillar cell contraction hasbeen shown to cause a more equal and better perfusion of lamellae(Sundin and Nilsson, 1998; Stenslokken et al., 2006). This probablywould result in an increase in respiratory surface area (Coolidgeet al., 2008). This effect, however, has not yet been demonstratedfor trout lamellae (Sundin and Nilsson, 1997).
Thus all studies consistently reported a branchial vasoconstric-tion following serotonin application, and in trout and Antarcticborch this vasoconstriction impaired branchial gas exchange. Thissuggests that the serotonergic influence on pillar cells may notbe very strong, because this usually would improve gas exchange.Although the slightly improved gas exchange observed in the Euro-pean eel following serotonin application is not fully explained yet,it may indicate that species specific differences in the serotoninresponsiveness of gill tissue exist.
4. Hypoxia and serotonergic signaling
A discharge of neural activity was recorded from afferent fibersof cranial nerves IX and X innervating isolated gill arches exposedto hypoxia in vitro (Burleson and Milsom, 1993; Milsom and Brill,1986) indicating that gill cells detect hypoxic conditions and thusmay initiate cardiovascular responses to reduced oxygen availabil-ity. Hypoxia in fact elicits a set of cardiovascular and respiratoryadaptations in fish (Mandic et al., 2009; Pelster and Bagatto,2010; Randall, 1982; Richards et al., 2009), and branchial vascu-lar resistance is known to increase (Holeton and Randall, 1967).Furthermore hypoxia has been reported to induce a degranula-tion of serotonin containing NECs (Dunel-Erb et al., 1982, 1989),and chronic hypoxia induced hypertrophy, proliferation and pro-cess extension in zebrafish NEC’s immunoresponsive to serotoninor synaptic vesicle protein (Jonz et al., 2004).
Based on these observations it was tempting to assume thatthe hypoxic branchial vasoconstriction did involve a serotoner-gic response and that the serotonin-mediated vasoconstriction of
the proximal vessels in the gills would cause an increase in theperfusion pressure and thus facilitate perfusion of the more dis-tal lamellae, resulting in a recruitment of lamellae (Bailly et al.,1989). However, the decrease in arterial PO2 reported for trout
3 siolog
ataiwcv(tabaccao
tuotacwvctesrgs
5a
wmaereedtbtcTbactt
anasp
NbF
44 B. Pelster, T. Schwerte / Respiratory Phy
nd borch as a consequence of serotonergic stimulation was coun-erintuitive (Fritsche et al., 1992; Sundin et al., 1998). A detailednalysis of the effect of hypoxia on gill blood flow revealed anncrease in the perfusion of the arterio-venous pathway in cod,
hich is supposed to bring oxygenated blood back to the coronaryirculation of the heart. This apparently can be achieved by an acti-ation of the sphincter associated with the efferent filament arterySundin, 1995). While an �-adrenergic vasoconstriction was iden-ified, it was hypothesized that the opening of the arterio-venousnastomosis guiding blood into the arterio-venous pathway mighte related to serotonergic effects. This hypothesis, however, is inpparent contrast to previous reports demonstrating that serotoninaused a vasoconstriction of branchial vessels. Another possibilityould be that the constriction of the efferent filament artery gener-tes the pressure to open the arterio-venous anastomoses and thuspens the arterio-venous pathway.
Careful microscopic studies of the microvascular blood flow inrout gills resulted in a more detailed picture of gill blood flownder hypoxic conditions (Sundin and Nilsson, 1997). As alreadybserved in cod hypoxia resulted in an increase in branchial resis-ance due to activation of the sphincter at the efferent filamentrtery, followed by a significant redirection of flow through theentral venous sinus. Blood velocity in the efferent filament arteryas severely reduced, occasionally blood flow even stopped in this
essel. Pharmacological treatment elucidated that this effect couldompletely be mimicked by cholinergic agonists, and preincuba-ion with methysergide to inhibit 5-HT receptors was without anyffect. The authors concluded that the hypoxic branchial vasocon-triction was a cholinergic reflex that did not include any otheregulatory mechanism (Sundin and Nilsson, 1997). Nevertheless,iven the complexity of the regulatory mechanisms there may bepecies specific differences.
. Serotonergic interaction with adrenergic, cholinergicnd purinergic signaling
Some of these serotonergic neurons are in synaptic contactsith catecholaminergic nerve fibers suggesting the existence of aodulatory relationship between the sympathetic and the cranial
utonomic nerves supplying the teleost gill (Bailly et al., 1989). Sev-ral studies suggested that serotonin may induce catecholamineelease in trout (Fritsche et al., 1993; Katchen et al., 1976). Strongvidence for a connection between the serotonergic and the adren-rgic pathway comes from the observed effect of serotonin onorsal arterial blood pressure. Studies consistently demonstratehat serotonergic stimulation results in a decrease in dorsal arteriallood pressure, but preincubation with the known 5-HT recep-or blockers cannot prevent this effect, while the branchial effectsan be totally abolished (Fritsche et al., 1992; Sundin et al., 1998).he decrease in dorsal arterial pressure therefore is assumed toe mediated by adrenergic signaling, provoked by serotonergicctivation. Serotonin may activate adrenergic neurons, which thanause the vasoconstriction. Because the typically used 5-HT recep-or blockers do not inhibit this effect, it must be assumed that inhese neurons a different 5-HT receptor type is involved.
The sphincter muscle associated with the efferent filamentrtery is innervated by cholinergic fibers, but there is also seroto-ergic stimulation (Bailly et al., 1989; Dunel-Erb et al., 1989; Sundinnd Nilsson, 2002; Jonz and Nurse, 2003). This dual responsivenessuggests that there may be interaction between these two signalingathways, although the nature of this interaction is not yet clear.
While serotonergic immunoreactivity is well established forECs, in addition, P2X3-IR receptors (= purinergic receptors) haveeen identified on NECs in close association with pillar cells.or NECs with purinergic responsiveness no innervation could be
y & Neurobiology 184 (2012) 340– 346
detected, supporting the possible paracrine role of these cells. Ittherefore has been speculated that there may also be an interactionbetween the serotonergic and the purinergic signaling pathways(Jonz and Nurse, 2003).
6. Selective serotonin reuptake inhibitors in theenvironment
Fluoxetine, the active compound in ProzacTM, is a selective sero-tonin reuptake inhibitor (SSRI) that prevents reuptake of serotoninin a variety of human cells. It is used most commonly for the treat-ment of illnesses such as depression and anxiety as well as panicand obsessive-compulsive disorders. Cell metabolism converts flu-oxetine to norfluoxetine. Norfluoxetine can also act as an SSRI withsimilar potency as fluoxetine. Fluoxetine and norfluoxetine areexcreted from the human body primarily through urine, with upto 11% of the administered fluoxetine dose being excreted (Hiemkeand Härtter, 2000).
Both compounds enter wastewater treatment facilities due tohuman excretion and through the disposal of unused drugs. As aconsequence, low levels of fluoxetine and its potent metabolitehave been continuously released into the environment, mean-while resulting in measurable quantities in freshwater (Hiemke andHärtter, 2000) and marine coastal ecosystems (Vasskog et al., 2008).
Because fluoxetine treatment results in an increase in 5-HT con-centrations, any physiological process that is regulated by 5-HTmay be susceptible to environmental fluoxetine exposure; par-ticularly those that are present in organs exposed directly to theenvironmental water. In the case of marine teleost fish, affectedorgans were shown to include the gills and the gastrointestinaltract (Morando et al., 2009). Fluoxetine exposure may lead toincreased plasma serotonin levels, and serotonin is a potent vaso-constrictor of the arterio-arterial branchial vasculature (see above).Consequently, decreased gill blood flow may lead to impaired gasexchange and systemic hypoxia.
Although to date there appears to be no systematic study char-acterizing the effect of water borne fluoxetine on gill perfusiondirectly, there is evidence for a systemic action of fluoxetine in fish.The majority of studies show behavioral effects (Airhart et al., 2007;Beulig and Fowler, 2008; Egan et al., 2009), an influence on repro-duction (Mennigen et al., 2008) and changes in osmoregulation andnitrogen waste excretion (Morando et al., 2009).
7. Conclusions
Several studies addressed the role of serotonergic signaling ingills, but at a closer look the information available so far remainsfragmentary and does not generate a conclusive picture. At theultrastructural level of serotonergic cells and with respect toserotonergic function several questions remain. Among the sero-tonergic cells so-called polymorphous granular cells have beendescribed (Dunel-Erb et al., 1982; Bailly et al., 1989, 1992; Zacconeet al., 1992), but these cells have not been mentioned in later stud-ies. It is tempting to assume that these polymorphous granularcells currently are considered to be NECs, but this is just a spec-ulation. Zaccone et al. (Zaccone et al., 1992) described open-typeand closed-type NECs. Closed-type NECs are located basally in thegill epithelium and have processes that never reach the surface ofthe epithelium, while open-type cells may reach from the base-ment membrane to the surface epithelium. Jonz and Nurse (Jonzand Nurse, 2003) mentioned these cells in their studies, but it is
not yet clear whether these cells do serve differential functions.Furthermore, the question whether non-serotonergic NECs are notyet fully differentiated cells remains unanswered (Jonz and Nurse,2003; Saltys et al., 2006).
siolog
etspsvb
treiedtrwobAlbo
A
m
R
A
A
B
B
B
B
B
B
C
D
D
D
D
E
F
flow by serotonin in the rainbow trout, Oncorhynchus mykiss. American Journal
B. Pelster, T. Schwerte / Respiratory Phy
Paracrine action of serotonin causes a vasoconstriction of thefferent filament artery; especially the sphincter associated withhis artery is responsive. It has not been proven, however, whethererotonin may have a vasodilatory effect on the arterio-venousathway in the gills, or whether the constriction of efferent ves-els indirectly opens the vessels of this pathway. The increase inentral aortic blood pressure following serotonin injection woulde in line with this assumption (Fritsche et al., 1992).
Serotonergic action has frequently been discussed with respecto hypoxia (Bailly et al., 1989). The decrease in arterial oxygen satu-ation following hypoxic stimulation (Fritsche et al., 1992; Sundint al., 1998) and the decrease in the perfusion of distal lamellaes clearly contradictive to this assumption. Furthermore, cholin-rgic branchial vasoconstriction in response to hypoxia has beenemonstrated, but serotonergic effects could not be shown underhese conditions. Serotonergic function has been discussed withespect to ion regulation, acid-base regulatory mechanisms andater balance (Fritsche et al., 1992; Sundin et al., 1998), but to
ur knowledge none of these environmental stress factors has evereen tested with respect to serotonergic function in gill perfusion.ccordingly, although serotonin is a powerful vasoconstrictor at
east for the efferent filament artery, no physiological situation haseen described to date in which serotonergic signaling is unequiv-cally involved in the control of gill perfusion.
cknowledgment
The authors would like to thank Michael Jonz for critical com-ents on the manuscript.
eferences
driaensen, D., Scheuermann, D.W., 1993. Neuroendocrine cells and nerves of thelung. Anatomical Record 236, 70–86.
irhart, M.J., Lee, D.H., Wilson, T.D., Miller, B.E., Miller, M.N., Skalko, R.G., 2007.Movement disorders and neurochemical changes in zebrafish larvae after bathexposure to fluoxetine (PROZAC). Neurotoxicology and Teratology 29, 652–664.
ailly, Y., Dunel-Erb, S., 1986. The sphincter of the efferent filament artery in teleostgills: 1. Structure and parasympathetic innervation. Journal of Morphology 187,219–237.
ailly, Y., Dunel-Erb, S., Geffard, M., Laurent, P., 1989. The vascular and epithe-lial serotonergic innervation of the actinopterygian gill filament with specialreference to the trout, Salmo gairdneri. Cell and Tissue Research 258, 349–363.
ailly, Y., Dunel-Erb, S., Laurent, P., 1992. The neuroepithelial cells of the fish gill fila-ment: indolamine-immunocytochemistry and innervation. Anatomical Record233, 143–161.
eulig, A., Fowler, J., 2008. Fish on prozac: effect of serotonin reuptake inhibitors oncognition in goldfish. Behavioral Neuroscience 122, 426–432.
urleson, M.L., Milsom, W.K., 1993. Sensory receptors in the first gill arch of rainbowtrout. Respiration Physiology 93, 97–110.
urleson, M.L., Smatresk, N.J., 2000. Branchial chemoreceptors mediate ventilatoryresponses to hypercapnic acidosis in channel catfish. Comparative Biochemistryand Physiology Part A: Molecular and Integrative Physiology 125, 403–414.
oolidge, E.H., Ciuhandu, C.S., Milsom, W.K., 2008. A comparative analysis of puta-tive oxygen-sensing cells in the fish gill. Journal of Experimental Biology 211,1231–1242.
unel-Erb, S., Bailly, Y., 1986. The sphincter of the efferent filament artery in teleostgills: II. Sympathetic innervation. Journal of Morphology 187, 239–246.
unel-Erb, S., Bailly, Y., Laurent, P., 1982. Neuroepithelial cells in fish gill primarylamellae. Journal of Applied Physiology 53, 1342–1353.
unel-Erb, S., Laurent, P., 1977. Physiologie comparée: la vascularisation branchialechez l‘anguille: Action de l‘acétylcholine et de adrénaline sur la répartition d‘uneresine polymérisable dans les différents compartiments vasculaires. ComptesRendus de l’Académie des Sciences (Paris) 284, 2011–2014.
unel-Erb, S., Bailly, Y., Laurent, P., 1989. Neurons controlling the gill vasculature infive species of teleosts. Cell and Tissue Research 255, 567–573.
gan, R.J., Bergner, C.L., Hart, P.C., Cachat, J.M., Canavello, P.R., Elegante, M.F.,Elkhayat, S.I., Bartels, B.K., Tien, A.K., Tien, D.H., Mohnot, S., Beeson, E., Glas-gow, E., Amri, H., Zukowska, Z., Kalueff, A.V., 2009. Understanding behavioraland physiological phenotypes of stress and anxiety in zebrafish. Behavioural
Brain Research 205, 38–44.
orster, M.E., Forster, A.H., Davison, W., 1998. Effects of serotonin, adrenaline andother vasoactive drugs on the branchial blood vessels of the Antarctic fishPagothenia borchgrevinki. Fish Physiology and Biochemistry 19, 103–109.
y & Neurobiology 184 (2012) 340– 346 345
Fritsche, R., Reid, S.G., Thomas, S., Perry, S.F., 1993. Serotonin-mediated release of cat-echolamines in the rainbow trout Oncorhynchus mykiss. Journal of ExperimentalBiology 178, 191–204.
Fritsche, R., Thomas, S., Perry, S.F., 1992. Effects of serotonin on circulation andrespiration in the rainbow trout Oncorhynchus mykiss. Journal of ExperimentalBiology 173, 59–73.
Hiemke, C., Härtter, S., 2000. Pharmacokinetics of selective serotonin reuptakeinhibitors. Pharmacology and Therapeutics 85, 11–28.
Holeton, G.F., Randall, D.J., 1967. Changes in blood pressure in the rainbow troutduring hypoxia. Journal of Experimental Biology 46, 297–305.
Janvier, J., 1997. Mediation of serotonin-induced branchial vasoconstriction by acholinergic and muscarinic response in vivo in European eel Anguilla anguilla.Fish Physiology and Biochemistry 16, 85–92.
Jonz, M.G., Fearon, I.M., Nurse, C.A., 2004. Neuroepithelial oxygen chemoreceptorsof the zebrafish gill. Journal of Physiology (London) 560, 737–752.
Jonz, M.G., Nurse, C.A., 2003. Neuroepithelial cells and associated innervation of thezebrafish gill: a confocal immunofluorescence study. Journal of ComparativeNeurology 461, 1–17.
Katchen, M.S., Olson, K.R., Wayne, C., 1976. Effects of histamine and serotonin onisolated perfused gill of rainbow trout (Salmo gairdneri). Federation Proceedings.Federation of American Socities for Experimental Biology 35, 528.
Mandic, M., Todgham, A.E., Richards, J.G., 2009. Mechanisms and evolution ofhypoxia tolerance in fish. Proceedings of the Royal Society B: Biological Science276, 735–744.
McKendry, J.E., Milsom, W.K., Perry, S.F., 2001. Branchial CO(2) receptors and car-diorespiratory adjustments during hypercarbia in Pacific spiny dogfish (Squalusacanthias). Journal of Experimental Biology 204, 1519–1527.
McKenzie, D.J., Burleson, M.L., Randall, D.J., 1991. The effects of branchial denerva-tion and pseudobranch ablation on cardioventilatory control in an air-breathingfish. Journal of Experimental Biology 161, 347–365.
Mennigen, J.A., Martyniuk, C.J., Crump, K., Xiong, H., Zhao, E., Popesku, J., Anis-man, H., Cossins, A.R., Xia, X., Trudeau, V.L., 2008. Effects of fluoxetine on thereproductive axis of female goldfish (Carassius auratus). Physiology General 35,273–282.
Milsom, W.K., Brill, R.W., 1986. Oxygen sensitive afferent information arising fromthe first gill arch of yellowfin tuna. Respiration Physiology 66, 193–203.
Morando, M.B., Medeiros, L.R., McDonald, M.D., 2009. Fluoxetine treatment affectsnitrogen waste excretion and osmoregulation in a marine teleost fish. AquaticToxicology 93, 253–260.
Nilsson, S., 1984. Adrenergic control systems in fish. Marine Biology Letters 5,127–146.
Nilsson, S., Pettersson, K., 1981. Sympathetic nervous control of blood flow in thegills of the Atlantic cod Gadus morhua. Journal of Comparative Physiology 144,157–163.
Olson, K.R., 1998. Hormone metabolism by the fish gill. Comparative Biochemistryand Physiology 119A, 55–65.
Olson, K.R., 2002. Vascular anatomy of the fish gill. Journal of Experimental Zoology293, 214–231.
Östlund, R., Fänge, R., 1962. Vasodilation by adrenaline and noradrenaline and theeffect of some other substances on perfused fish gill. Comparative Biochemistryand Physiology 5, 307–309.
Pelster, B., Bagatto, B., 2010. Respiration. In: Perry, S.F., Ekker, M., Farrell, A.P.,Brauner, C. (Eds.), Zebrafish. Academic Press, Amsterdam, pp. 290–309.
Pettersson, K., Nilsson, S., 1979. Nervous control of the branchial vascular resistanceof the Atlantic cod. Journal of Comparative Physiology 129, 179–183.
Randall, D., 1982. The control of respiration and circulation in fish during exerciseand hypoxia. Journal of Experimental Biology 100, 275–288.
Reite, O.B., 1969. The evolution of vascular smooth muscl response to histamine and5-hydroxytryptamine. I. Occurence of stimulatory actions in fish. Acta Physio-logica Scandinavica 75, 221–239.
Richards, J.G., Farrell, A.P., Brauner, C.J., 2009. Hypoxia. Fish Physiology 27.Saltys, H., Jonz, M., Nurse, C., 2006. Comparative study of gill neuroepithelial cells
and their innervation in teleosts and Xenopus tadpoles. Cell and Tissue Research323, 1–10.
Smith, D.G., 1977. Sites of cholinergic vasoconstriction in trout gills. American Jour-nal of Physiology – Regulatory, Integrative and Comparative Physiology 233,R222–R229.
Stenslokken, K.O., Sundin, L., Nilsson, G.E., 2006. Endothelin receptors in teleostfishes: cardiovascular effects and branchial distribution. American Journalof Physiology – Regulatory, Integrative and Comparative Physiology 290,R852–R860.
Sundin, L., Davison, W., Forster, M., Axelsson, M., 1998. A role of 5-HT2 receptorsin the gill vasculature of the Antarcitc fish Pagothenia borchgrevinki. Journal ofExperimental Biology 201, 2129–2138.
Sundin, L., Nilsson, G.E., 1997. Neurochemical mechanisms behind gill microcir-culatory responses to hypoxia in trout: in vivo microscopy study. AmericanJournal of Physiology – Regulatory, Integrative and Comparative Physiology 272,R576–R585.
Sundin, L., Nilsson, G.E., Block, M., Löfman, C.O., 1995. Control of gill filament blood
of Physiology 268, R1224–R1229.Sundin, L., Nilsson, G.E., 1998. Endothelin redistributes blood flow through the
lamellae of rainbow trout gills. Journal of Comparative Physiology B 168,619–623.
3 siolog
S
S
S
T
V
46 B. Pelster, T. Schwerte / Respiratory Phy
undin, L., Nilsson, S., 1992. Arterio-venous branchial blood flow in the Atlantic codGadus morhua. Journal of Experimental Biology 165, 73–84.
undin, L.I., 1995. Responses of the branchial circulation to hypoxia in the Atlanticcod, Gadus morhua. American Journal of Physiology – Regulatory, Integrativeand Comparative Physiology 268, R771–R778.
zaneva, V., Perry, S.F., 2010. The control of breathing in goldfish (Carassius aura-tus) experiencing thermally induced gill remodelling. Journal of Experimental
Biology 213, 3666–3675.
asskog, T., Anderssen, T., Pedersen-Bjergaard, S., Kallenborn, R., Jensen, E., 2008.Occurrence of selective serotonin reuptake inhibitors in sewage and receiv-ing waters at Spitsbergen and in Norway. Journal of Chromatography A 1185,194–205.
y & Neurobiology 184 (2012) 340– 346
Wilson, J.M., Laurent, P., 2002. Fish gill morphology: inside out. Journal of Experi-mental Zoology 293, 192–213.
Zaccone, G., Lauweryns, J.M., Fasulo, S., Tagliafierro, G., Ainis, L., Licata, A., 1992.Immunocytochemical localization of serotonin and neuropeptides in the neu-roendocrine paraneurons of teleost and lungfish gills. Acta Zoologica 73,177–183.
Zaccone, G., Mauceri, A., Fasulo, S., 2006. Neuropeptides and nitric oxide synthase inthe gill and the air-breathing organs of fishes. Journal of Experimental Zoology305A, 428–439.
Zaccone, G., Mauceri, A., Fasulo, S., Ainis, L., Lo Cascio, P., Ricca, M.B., 1996. Local-ization of immunoreactive endothelin in the neuroendocrine cells of fish gill.Neuropeptides 30, 53–57.
Zachar, P.C., Jonz, M.G., 2012. Confocal imaging of Merkel-like basal cells in the tastebuds of zebrafish. Acta Histochemica 114, 101–115.
