MINIREVIEW Osmotic stress sensing and signaling in fishes Diego F. Fiol and Dietmar Ku ¨ ltz Physiological Genomics Group, Department of Animal Science, University of California, Davis, CA, USA Physiological significance of osmotic stress for fishes Fishes represent the most ancient of five vertebrate classes. They originated more than 500 million years ago and have diverged into three major taxa: (a) hag- fishes and lampreys (Agnatha); (b) cartilagenous fishes (Chondrichthyii); and (c) ray-finned fishes (Actino- pterygii). These three taxa employ different strategies of systemic osmoregulation with only ray-finned fishes being strong osmoregulators. Nevertheless, at the cellu- lar level, all fish taxa (like other organisms) ionoregu- late to maintain K + and other intracellular inorganic ion concentrations within a tightly regulated range, which is essential to support cell metabolism. Like other aquatic (or semiaquatic) vertebrates (e.g. amphibians, alligators), fish are in direct contact with environmental water. Most fishes depend on stable water salinity to be able to osmoregulate and maintain constant osmolality in their body fluids (internal milieu). These are stenohaline species that can only live in either freshwater or seawater. Nonetheless, there are also numerous fish species that tolerate and even thrive in water characterized by greatly fluctuating salinity. Keywords euryhaline fishes; osmoregulation; osmosensing; osmotic stress; salinity adaptation; stress signaling Correspondence D. Ku ¨ ltz, Comparative Physiological Genomics Group, Department of Animal Science, One Shields Avenue, Meyer Hall, University of California, Davis, CA 95616, USA Fax: +1 530 752 0175 Tel: +1 530 752 2991 E-mail: dkueltz@ucdavis.edu (Received 2 July 2007, accepted 7 September 2007) doi:10.1111/j.1742-4658.2007.06099.x In their aqueous habitats, fish are exposed to a wide range of osmotic con- ditions and differ in their abilities to respond adaptively to these variations in salinity. Fish species that inhabit environments characterized by signifi- cant salinity fluctuation (intertidal zone, estuaries, salt lakes, etc.) are eury- haline and able to adapt to osmotic stress. Adaptive and acclimatory responses of fish to salinity stress are based on efficient mechanisms of osmosensing and osmotic stress signaling. Multiple osmosensors, including calcium sensing receptor likely act in concert to convey information about osmolality changes to downstream signaling and effector mechanisms. The osmosensory signal transduction network in fishes is complex and includes calcium, mitogen-activated protein kinase, 14-3-3 and macromolecular damage activated signaling pathways. This network controls, among other targets, osmosensitive transcription factors such as tonicity response ele- ment binding protein and osmotic stress transcription factor 1, which, in turn, regulate the expression of genes involved in osmotic stress acclima- tion. In addition to intracellular signaling mechanisms, the systemic response to osmotic stress in euryhaline fish is coordinated via hormone- and paracrine factor-mediated extracellular signaling. Overall, current insight into osmosensing and osmotic stress-induced signal transduction in fishes is limited. However, euryhaline fish species represent excellent models for answering critical emerging questions in this field and for elucidating the underlying molecular mechanisms of osmosensory signal transduction. Abbreviations CaSR, calcium sensing receptor; IEG, immediate early gene; MAPK, mitogen-activated protein kinase; Ostf1, osmotic stress transcription factor 1; TonEBP, tonicity response element binding protein; TRP, transient receptor potential. 5790 FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS Some of these euryhaline fish (e.g. tilapia) are able to live in freshwater as well as in water with salinities up to four times that of seawater. Thus, euryhaline fishes are able to inhabit environments characterized by severe osmotic stress, such as desert lakes, tidepools and estuaries. Euryhaline fishes have evolved physiological mecha- nisms that allow them to compensate the osmotic stress associated with fluctuating environmental salin- ity. An integral part of such physiological mechanisms is the ability to sense and quantify changes in environ- mental salinity and to activate appropriate compensa- tory responses. Thus, euryhaline fishes represent excellent models to identify and understand elements and mechanisms controlling the physiological and behavioral changes that occur in response to osmotic stress. The three major groups of players involved in this response are osmosensors, signal transducers and effectors. Osmosensors control signal transduction net- works that, in turn, regulate effector mechanisms responsible for acclimation to changes in environmen- tal salinity (Fig. 1). Many effector mechanisms involved in osmotic acclimation of euryhaline fishes have been identified and characterized in detail but little is known about the proximal osmosensors and signal transduction pathways that control these effec- tor mechanisms. In what follows, we will briefly review our current knowledge about osmosensory signal transduction in euryhaline fishes and compare it with knowledge available for some other animals. Osmotic stress sensing in fishes General considerations about molecular osmosensors Our knowledge of molecular osmosensors that monitor and quantify environmental and extracellular osmolal- ity in fishes is minimal. As is true for other cells, it is not clear how fish cells quantify osmolality to mount a compensatory adaptive response of proper magnitude or, alternatively, induce programmed cell death when their tolerance limit is exceeded. Many different types of molecular osmosensors can be envisioned because osmotic stress impacts essentially all cellular structures, processes and macromolecules. Potential osmosensors include membrane proteins that are regulated by ion concentration or membrane stretching and compac- tion, molecular chaperones that monitor the degree of protein unfolding, DNA damage sensors, proteins associated with cytoskeletal organization, and enzymes whose activity is stringently correlated with intracellu- lar electrolyte concentration. It is likely that multiple molecular osmosensors act in concert to control osmosensory signal transduction networks and that some of them are activated more over a range of mild osmotic stress whereas others are activated more over a range of severe osmotic stress. In addition, many proximal events perceived by poten- tial osmosensors (cell volume changes, changes in cytoskeletal organization, membrane stretching or compaction, molecular crowding) are only prevalent during more severe and acute osmotic stress. Such events result from osmosis across semipermeable mem- branes of animal cells but osmosis may not occur when osmolality changes happen gradually over an extended period of time. Under these conditions, it is most likely intracellular ionic strength that serves as the initial signal triggering molecular osmosensors. Equilibration of intracellular ionic strength during gradual osmolality changes can be achieved without osmosis if the capacity of ion transport proteins in cell membranes is sufficient for moving ions across mem- branes at a rate that offsets water movement across membranes. This is only possible for small and slow osmolality changes and it depends on cell type-specific Fig. 1. Major elements of the osmosensory signal transduction network in fishes. Multi- ple osmosensors (see text) recognize osmolality ⁄ salinity changes and activate a signaling network that integrates the infor- mation received from different osmosen- sors, amplifies this information, and turns on ⁄ off a large number of appropriate effec- tor mechanisms (i.e. mechanisms of physio- logical acclimation). D. F. Fiol and D. Ku ¨ ltz Osmotic stress sensing and signaling in fishes FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS 5791 composition of the cell membrane, including the abun- dance of particular ion channels, ion transport pro- teins, aquaporins and membrane lipids. Thus, different cell types within the same organism may be able to sense different ranges of osmolality changes. This abil- ity is critical for euryhaline fishes and other aquatic vertebrates because some of their cells (e.g. gill cells) are exposed directly to the aquatic environment and experience very wide ranges of osmolality whereas most other cells are bathed in a more homeostatic environment as a result of systemic osmoregulation. Changes in extracellular fluid osmolality (i.e. plasma osmolality) in aquatic vertebrates such as fishes are also sensed via perception of concomitant fluid volume and ⁄ or blood pressure changes. Systemic osmosensors and baroreceptors are responsible for monitoring plasma osmolality and they are conserved in all verte- brates. Peripheral systemic osmosensors of fishes appear to be located primarily in the gills [1] and pitui- tary gland [2]. Once triggered, molecular and systemic osmosensors activate a signaling network that, in turn, controls effector mechanisms mediating physiological acclimation to osmotic stress. Putative molecular osmosensors in fish cells Molecular osmosensors of fish cells are not well char- acterized. However, analysis of zebrafish and pufferfish genomes shows that putative molecular osmosensors of mammalian and invertebrate cells are highly conserved in fish genomes (Fig. 2). Such putative osmosensors include adenyl cyclase [3], transient receptor potential (TRP) channels [4], and aquaporin 4 [5]. However, functional evidence firmly establishing these proteins as molecular osmosensors in euryhaline fishes is lack- ing and this area needs to be experimentally addressed in future research. A role of adenyl cyclase as an osmosensor in eury- haline fishes is supported by its effects on chloride secretion across the gill epithelium and osmoregulatory hormone secretion from the pituitary gland. Forskolin, which stimulates adenyl cyclase activity, was shown to enhance chloride secretion across opercular membranes of euryhaline fishes [6], as well as prolactin and growth hormone secretion from trout pituitary gland [7]. These secretory processes are also stimulated when euryhaline fish face salinity increases. Nevertheless, it is not known whether activation of adenyl cyclase in euryhaline fishes is directly mediated by osmolality changes as would be required for a true osmosensor protein. Osmosensory TRPV4 channels were localized in Danio rerio and the expression of this channel protein in the developing kidney was demonstrated [8]. More- over, evolutionary studies on the TRP protein family identified six copies of TRPV4 in the western clawed frog (Xenopus tropicalis), suggesting that diversification of osmosensory TRPV4 may favor adaptation to both aquatic and terrestrial environments, which represent very different habitats regarding requirements for osmoregulation [9]. However, as for adenyl cyclase, no direct evidence for an osmosensory function of TRP channels in fishes has been published. Aquaporin water channels have been studied in fishes, including their regulation during salinity stress. Nevertheless, all studies to date have focused on the role of aquaporins as effector proteins of osmosensory signal transduction pathways and the potential role of these proteins in osmosensing of fish cells has yet to be addressed. As a result of recent studies on fish aqu- aporins, we know that changes in water permeability in gills and intestine are mediated at least in part via regulation of aquaporin abundance in epithelial cell membranes [10]. Thus, a role of aquaporins as impor- tant effector proteins of osmoregulation in fishes has been established. It will be interesting to see whether water channels also function as systemic osmosensors in the brain of fishes, as has been suggested for mammals [5]. Calcium sensing receptor The calcium sensing receptor (CaSR) has been identi- fied as an important osmosensor protein in fishes. CaSR is a large glycoprotein belonging to the G pro- tein-coupled receptor superfamily. This membrane pro- tein is regulated directly by extracellular calcium (and to some extent also other polyvalent cations) as ligand in the millimolar range. Fishes (e.g. euryhaline marine species) utilize CaSR for sensing environmental salinity [11]. In particular for marine fishes, the calcium con- centration in the external environment (seawater) is in the millimolar range that is accurately sensed by CaSR. Thus, changes in environmental calcium con- centration are thought to be a surrogate measure for the ionic strength ⁄ salinity of the marine environment. In agreement with this notion, CaSR is expressed in osmoregulatory tissues of fishes, including shark rectal gland [12] and teleost gill and opercular membrane [13]. Full-length transcripts of CaSR have been cloned from gilthead sea bream (Sparus aurata) [13] and spiny dogfish (Squalus acanthias) [11]. Using nucleotide probes, CaSR transcripts have been localized to bran- chial chloride cells of both aforementioned species, as well as winter flounder (Pleuronectes americanus) and Atlantic salmon (Salmo salar) [11]. Osmotic stress sensing and signaling in fishes D. F. Fiol and D. Ku ¨ ltz 5792 FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS Tilapia CaSR senses changes in external [Ca 2+ ] and activates phospholipase C and mitogen-activated pro- tein kinase (MAPK) signaling [14]. Moreover, changes in plasma [Ca 2+ ] and [Mg 2+ ] that occur when fish move from freshwater to seawater, or vice versa, likely serve as salinity sensing cues for CaSR because plasma 4e-19 5e-32 2e-18 4e-35 9e-18 3e-35 RHD IPT_NFAT 1e-08 2e-15 3e-15 TSC22 1e-53 3e-45 1e-52 M/P 3e-10 4e-10 1e-12 7e-10 3e-12 4e-10 ANK ION_TRANS Fig. 2. Evolutionary conservation of orthologs of the putative osmosensors TRPV4 and aquaporin 4 and the osmosensory signal transcription proteins TonEBP and TSC22D2 in vertebrates. The human sequences of TRPV4 (871 amino acids, AAG28029.1), aquaporin 4 (323 amino acids, NP_001641.1), TonEBP (1531 amino acids, NP_006590.1), and the mouse sequence of TSC22D2-4 (116 amino acids, EU004151) were used as references and their conserved domains are indicated. The highest homology hits for each D. rerio, X. tropicalis and Gallus gallus ge- nomes were analyzed for the presence of conserved domains in the Conserved Domain Database and Search Service, version 2.11 (NCBI, 17402 motifs) and the expectation values are indicated. Percentages of amino acid sequence similarity and identity are shown. RHD, Rel homology domain (pfam00554); IPT_NFAT, IPT domain of the NFAT family of transcription factors (cd01178); TSC22, TSC-22 ⁄ dip ⁄ bun family (pfam01166); MIP, major intrinsic protein (cd00333); ANK, ankyrin repeats (cd00204); ION_TRANS, ion transport protein (pfam00520). D. F. Fiol and D. Ku ¨ ltz Osmotic stress sensing and signaling in fishes FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS 5793 concentrations of these divalent ions are well within the millimolar range over which CaSR operates [11]. In the mammalian kidney, CaSR regulates the activity of many other signaling pathways, including pathways that are regulated by intracellular calcium concentra- tion. The evidence briefly summarized above suggests that CaSR plays a significant role for osmosensing in fishes. Osmosensory signal transduction network in fishes Studies on bacteria, yeast and model animals have shown that osmosensors control an elaborate intracel- lular signaling network. The major role of this network is to integrate signals from multiple osmosensors and generate an amplified output-stimulus for controlling appropriate effector mechanisms (Fig. 1). We hypothe- size that the mode of integration of signals generated by multiple osmosensors with different sensitivity ranges enables cells to determine the severity of osmo- tic stress, quantify extracellular osmolality, and ensure that an appropriate physiological response is mounted. Testing this hypothesis will require detailed knowledge about the key elements involved in osmosensory signal transduction. Known elements of osmosensory signal transduction in euryhaline fishes are calcium-dependent pathways, MAPKs, 14-3-3 proteins, specific transcrip- tion factors, hormones, and paracrine factors. Their role during osmotic stress is briefly reviewed below. Role of intracellular calcium We have summarized above that environmental cal- cium may be an important trigger of osmosensory events by controlling CaSR activity. In addition, many effects of changes in environmental and plasma cal- cium concentration on fish gill chloride cell morphol- ogy and the function of important osmoregulatory effector proteins have been documented [15]. Since cal- cium is a major second messenger in eukaryotic cells and known to play significant roles in osmosensory signal transduction of mammalian and even plant cells, it is very likely that calcium-mediated signaling con- tributes significantly to osmosensory signal transduc- tion in fish cells. The importance of intracellular calcium for the activation of downstream signaling events in fish exposed to osmotic stress has been stud- ied in fish rostral pars distalis cells. These cells are excellent models because they represent a relatively homogeneous (approximately 97%) population of pro- lactin secreting cells and their prolactin secretion depends on osmolality. In tilapia, hyposmotic stress stimulates prolactin secretion, which was shown to depend on stretch-activated ion channels and increased intracellular calcium [16]. Cortisol, a hormone associ- ated with hyperosmotic stress, inhibits prolactin secre- tion via reduction of free intracellular calcium [17]. In addition to its effect on intracellular calcium, cortisol also inhibits adenyl cyclase, a potential osmosensor mentioned above, suggesting that both major intra- cellular second messengers, calcium and cAMP, are involved in osmotic stress signaling [18]. Another osmoregulatory hormone, angiotensin II, increases free intracellular calcium in fish tissues [19], confirming that the effects of osmoregulatory hormones are mediated at least partly via intracellular calcium signaling. An important role of intracellular calcium in fish osmotic stress signaling is also supported by a modeling approach yielding an osmosensory signal transduction network based on 20 immediate early genes that rap- idly respond to salinity stress in tilapia gill. Intracellu- lar calcium is a major node in this network, which also contains several calcium-binding proteins such as an- nexins and S-100 proteins [20]. Notably, annexins and two other immediate early genes (IEGs) identified in this study (gelsolin, galectin 4) are known to regulate actin-based cytoskeleton remodeling in mammalian cells, suggesting that this process may be a major tar- get during osmotic stress acclimation in fish gill cells. Consistent with this view, the actin-based cytoskeleton seems to play a role in osmotic regulation of Na + ⁄ K + ⁄ 2Cl – (NKCC) cotransporter [21] and in the closing or opening of apical crypts of gill chloride cells [22]. Furthermore, changes in ion transport during hyper- and hypotonic stress require intact F-actin and microtubules in eel intestinal epithelium [23]. MAPK MAPKs are a family of enzymes that are involved in osmosensory signal transduction in yeast, plant and animal cells. They are key elements of protein phos- phorylation cascades that integrate and amplify signals from osmosensors to activate appropriate downstream targets mediating physiological acclimation. Although MAPKs are highly evolutionarily conserved, their acti- vators and substrates can differ greatly, depending on taxon, physiological condition and developmental state. For example, yeast exposed to osmotic stress activate the high osmolarity glycerol response (HOG1) MAPK cascade via the SLN1 osmosensor, which is a two-component histidine kinase, none of whose com- ponents are present in any sequenced animal genome. This illustrates that osmotic stress signaling networks are modular. Recent evidence suggests that MAPK Osmotic stress sensing and signaling in fishes D. F. Fiol and D. Ku ¨ ltz 5794 FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS cascades represent an important module of such net- works in euryhaline fish. We have shown that the activity ⁄ phosphorylation of all three major MAPKs is rapidly altered in gill epithelium of killifish (Fundulus heteroclitus) when these fish experience osmotic stress in vivo [20]. Osmotic regulation of p38 MAPK and JNK (Jun-N-terminal kinase ⁄ stress-activated protein kinase) MAPK phosphorylation was also observed in isolated opercular epithelium of killifish, where chlo- ride secretion decreases after addition of a pharmaco- logical p38 inhibitor [24]. Furthermore, p38 MAPK is required for regulatory volume decrease in isolated hepatocytes from turbot (Scophthalmus maximus) [25]. We recently identified an upstream regulator of MAPK cascades, mitogen-activated protein kinase kinase kinase 7 interacting protein 2 (TAK 1 binding protein 2 ¼ TAB 2), as an IEG during hyperosmotic stress in tilapia gill epithelium [26]. This gene is transiently and very rapidly (within 2 h) induced by hyperosmotic stress, indicating a role of this mitogen- activated protein kinase kinase kinase 7 interacting protein for osmosensory signal transduction in fish. 14-3-3 proteins 14-3-3 proteins are evolutionarily highly conserved in all eukaryotes. They sequester other proteins that are phosphorylated on serine (and sometimes threonine). They operate as dimers with each monomer binding one phosphoprotein. Thus, 14-3-3 proteins can be con- sidered nodes that bring together elements of phos- phorylation-based signal transduction networks. In addition, they promote subcellular translocation of phosphoproteins (e.g. from nucleus into cytosol) and thereby affect the function of phosphoproteins. We have cloned the first fish 14-3-3 protein from the eury- haline killifish (F. heteroclitus) and shown that its abundance is regulated in gill epithelium by environ- mental salinity [27]. Surprisingly, osmotic 14-3-3 regu- lation is very slow in this fish (it takes many hours) and we reason that regulation of 14-3-3 abundance may represent a secondary response. Rapid regulation of 14-3-3 binding to phospho-proteins may be medi- ated by post-translational modification or dimerization but this remains to be investigated. Of interest, hetero- logous expression of F. heteroclitus 14-3-3 in Xenopus laevis oocytes protects the oocytes from osmotic stress, which was attributed to its inhibition of an endoge- nous oocyte chloride current [28]. 14-3-3 proteins are strategically positioned at points of cross-talk between virtually all important cell signaling pathways. There- fore, identification of 14-3-3 binding partners during salinity acclimation of fish should provide a new window into osmosensory signal transduction mecha- nisms. Transcription factors Many physiological acclimations to environmental changes are mediated by alteration of gene expression and there are numerous studies thoroughly validating the critical importance of this mechanism for osmotic stress acclimation in euryhaline fish. Thus, inducible transcription factors contributing to changes in gene expression during osmotic stress are of great interest. In mammals, the tonicity response element binding protein (TonEBP) transcription factor (NFAT5, ORE- BP) plays a major role in response to hypertonicity [29]. Sequence similarity searches identified genes encoding orthologous proteins in the fishes D. rerio, Takifugu rubripes and Tetraodon nigroviridis (Fig. 2). The presence of TonEBP in fish genomes raises the possibility that it plays a role for osmotic stress signal- ing in fishes. Indeed, a recent study on killifish (F. het- eroclitus) provides experimental evidence that TonEBP participates in osmosensory signal transduction in fish cells [30]. Recently, we identified two putative transcriptional regulators, osmotic stress transcription factor 1 (Ostf1) and basal transcription factor IIB, as early hyper- osmoticaly up-regulated proteins in tilapia gills [31]. We demonstrated that Ostf1 up-regulation depends on RNA stabilization and transcriptional mechanisms and on the presence of an osmotic gradient between the extracellular and intracellular fluid of tilapia gill cells [32]. A role of Ostf1 in osmosensory signal transduc- tion is not limited to fishes, but also is evident in mam- malian cells. We identified TSC22D2 as an Ostf1 ortholog of mammals and showed that it is activated and alternatively spliced in response to hypertonicity in mouse and human kidney cells with very similar kinetics as in fish gill cells. In addition, overexpression of mammalian TSC22D2 confers increased osmo- tolerance to murine inner medullary collecting duct cells [33]. Macromolecular damage response pathways Damage to macromolecules such as DNA and proteins represents an important sensory element for cellular recognition of severe environmental stress, including osmotic stress [34]. Thus, it is not surprising that among IEGs induced rapidly in response to hyper- osmotic stress in euryhaline tilapia are genes that rec- ognize macromolecular damage. Ubiquitin E3 ligase and the mRNA stabilizer HuR are particularly D. F. Fiol and D. Ku ¨ ltz Osmotic stress sensing and signaling in fishes FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS 5795 noteworthy in this context. In fish, two different ubiquitin E3 ligases are rapidly up-regulated during hyperosmotic stress, a Rbx1 homolog in salmon [35] and a Grail ⁄ Goliath homolog in tilapia [26], suggesting that they are fundamentally important in the osmotic stress response of fish. Ubiquitin E3 ligase may sense protein damage by quantifying the amount of sub- strates that it tags with ubiquitin. In most cases, such protein substrates are terminally damaged and destined for proteolytic degradation and removal. The under- lying molecular mechanisms by which ubiquitin E3 ligase keeps track of the amount of substrates it tags with ubiquitin and relays this information to the osmo- tic stress signaling network are not known. However, it is becoming increasingly clear that rapid and specific protein degradation via the ubiquitin-proteasome sys- tem is intrinsically linked to the regulation of adaptive gene expression, the cell cycle and adaptive cell differ- entiation. Moreover, in mammalian kidneys, the inter- action of Nedd4 E3 ubiquitin ligase with epithelial sodium channels is controlled by osmolality, vasopres- sin and 14-3-3 proteins [36]. Another potential ‘damage sensory’ protein that is rapidly induced by hyperosmotic stress in tilapia gills and stabilizes select mRNAs of adaptive value is HuR [26]. It is possible that HuR relays information about mRNA stability during hyperosmolality to the osmo- sensory signal transduction network. HuR recognizes, binds and, in most cases, stabilizes labile mRNAs. Thus, information about mRNA stability may contrib- ute to osmosensory signal transduction. Unfortunately, we know nothing about the role of the DNA damage sensing network for osmotic stress signaling in fish, although it is likely that information about DNA dam- age contributes to osmosensory signal transduction in fishes, just as it does in mammals and organisms other than vertebrates [34]. Systemic responses of fishes to osmotic stress Osmotic stress activates a systemic response, which is mediated by hormones to a great extent. Through their concerted action, osmoregulatory hormones coordinate adaptive responses in different tissues within an organ- ism [37]. Endocrine responses to osmotic stress seem to occur in two phases, an acute-phase response and a longterm response. The acute-phase response takes place in the order of minutes to hours and involves many hormones, including arginine vasopressin, angio- tensin II, natriuretic peptides, vasoactive intestinal pep- tide, urotensin II, insulin and nongenomic actions of corticosteroids [38]. Major outcomes of the acute- phase response are rapid changes in behavior, altered blood flow to osmoregulatory organs, alteration of membrane insertion of nascent ion transport proteins and changes in activity of existing ion transport pro- teins. The second phase long-term response is regulated primarily by genomic actions of corticosteroids, pro- lactin, growth hormone, and insulin-like growth fac- tor I. The effect of these hormones is fine-tuned at the cellular level via adjustment of expression and membrane insertion of the corresponding hormone receptors [39]. Signaling pathways emanating from those receptors control primarily long-term changes in ion transport capacity via regulation of transport pro- tein expression and synthesis, cell proliferation and cell differentiation [38]. Many excellent reviews, some of which are referenced above, have been published that detail the systemic action of osmoregulatory hor- mones in fishes, which exceeds the scope of this mini- review. Additional systemic factors contribute to osmoregu- lation in fishes via paracrine signaling. Such factors include endothelin, nitric oxide and prostanoids, which play a role in adaptive modulation of ion transport across the opercular epithelium of euryha- line killifish (F. heteroclitus) [40]. Nitric oxide, in par- ticular, may be a paracrine signal that contributes significantly to the regulation of chloride cell function in fish gills in response to osmotic stress because nitric oxide synthase is highly expressed in epithelial cells that are located immediately adjacent to chloride cells [41]. In summary, we have provided a brief overview about recent progress on osmosensing and osmotic stress signaling in fishes. Our current knowledge in this field is fragmentary at best and many interesting chal- lenges remain. Fundamental questions in this area still require answers. How do cells and organisms quantify osmotic stress to determine whether it exceeds their tolerance limits and adaptive capacity (e.g. whether or not to activate apoptosis)? How do key proteins and molecular mechanisms cooperate to confer high toler- ance to osmotic stress? Is the physiological trait of euryhalinity always based on the same conserved set of proteins and signaling mechanisms or did nature evolve many solutions to a common problem? How are osmotic stress signaling pathways integrated into intracellular signaling networks that control other physiological processes? Euryhaline fishes represent an excellent model for addressing those and many other questions that are emerging in this field because the relevant underlying mechanisms have evolved to great perfection in these animals. Osmotic stress sensing and signaling in fishes D. F. Fiol and D. Ku ¨ ltz 5796 FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS Acknowledgements This work was supported by grants from the National Science Foundation (IOB-0542755) and CALFED (SPSP2006-1035). References 1 Sundin L, Turesson J & Taylor EW (2003) Evidence for glutamatergic mechanisms in the vagal sensory pathway initiating cardiorespiratory reflexes in the shorthorn scul- pin Myoxocephalus scorpius. J Exp Biol 206, 867–876. 2 Seale AP, Fiess JC, Hirano T, Cooke IM & Grau EG (2006) Disparate release of prolactin and growth hor- mone from the tilapia pituitary in response to osmotic stimulation. Gen Comp Endocrinol 145, 222–231. 3 Saran S & Schaap P (2004) Adenylyl cyclase G is acti- vated by an intramolecular osmosensor. Mol Biol Cell 15, 1479–1486. 4 Liedtke W, Choe Y, Marti-Renom MA, Bell AM, Denis CS, Sali A, Hudspeth AJ, Friedman JM & Heller S (2000) Vanilloid receptor-related osmotically activated channel (VR-OAC), a candidate vertebrate osmorecep- tor. Cell 103, 525–535. 5 Venero JL, Vizuete ML, Ilundain AA, Machado A, Echevarria M & Cano J (1999) Detailed localization of aquaporin-4 messenger RNA in the CNS: preferential expression in periventricular organs. Neuroscience 94, 239–250. 6 May SA & Degnan KJ (1984) cAMP-mediated regula- tion of chloride secretion by the opercular epithelium. Am J Physiol 246, R741–R746. 7 Falcon J, Besseau L, Fazzari D, Attia J, Gaildrat P, Beauchaud M & Boeuf G (2003) Melatonin modulates secretion of growth hormone and prolactin by trout pituitary glands and cells in culture. Endocrinology 144, 4648–4658. 8 Mangos S, Liu Y & Drummond IA (2007) Dynamic expression of the osmosensory channel trpv4 in multiple developing organs in zebrafish. Gene Expr Patterns 7, 480–484. 9 Saito S & Shingai R (2006) Evolution of thermoTRP ion channel homologs in vertebrates. Physiol Genomics 27, 219–230. 10 Cutler CP, Martinez AS & Cramb G (2007) The role of aquaporin 3 in teleost fish. Comp Biochem Physiol A Mol Integr Physiol 148, 82–91. 11 Nearing J, Betka M, Quinn S, Hentschel H, Elger M, Baum M, Bai M, Chattopadyhay N, Brown EM, He- bert SC et al. (2002) Polyvalent cation receptor proteins (CaRs) are salinity sensors in fish. Proc Natl Acad Sci USA 99, 9231–9236. 12 Fellner SK & Parker L (2004) Ionic strength and the polyvalent cation receptor of shark rectal gland and artery. J Exp Zool Part A Comp Exp Biol 301A, 235– 239. 13 Flanagan JA, Bendell LA, Guerreiro PM, Clark MS, Power DM, Canario AVM, Brown BL & Ingleton PM (2002) Cloning of the cDNA for the putative calcium- sensing receptor and its tissue distribution in sea bream (Sparus aurata). Gen Comp Endocrinol 127, 117–127. 14 Loretz CA, Pollina C, Hyodo S, Takei Y, Chang W & Shoback D (2004) cDNA cloning and functional expres- sion of a Ca 2+ -sensing receptor with truncated C-termi- nal tail from the Mozambique tilapia (Oreochromis mossambicus). J Biol Chem 279, 53288–53297. 15 Wendelaar Bonga SE (1997) The stress response in fish. Physiol Rev 77, 591–625. 16 Seale AP, Richman NH, Hirano T, Cooke I & Grau EG (2003) Evidence that signal transduction for osmo- reception is mediated by stretch-activated ion channels in tilapia. Am J Physiol Cell Physiol 284, C1290–C1296. 17 Hyde GN, Seale AP, Grau EG & Borski RJ (2004) Cor- tisol rapidly suppresses intracellular calcium and volt- age-gated calcium channel activity in prolactin cells of the tilapia (Oreochromis mossambicus). Am J Physiol Endocrinol Metab 286, E626–E633. 18 Borski RJ, Hyde GN & Fruchtman S (2002) Signal transduction mechanisms mediating rapid, nongenomic effects of cortisol on prolactin release. Steroids 67, 539– 548. 19 Russell MJ, Klemmer AM & Olson KR (2001) Angio- tensin signaling and receptor types in teleost fish. Comp Biochem Physiol A Mol Integr Physiol 128, 41–51. 20 Ku ¨ ltz D & Avila K (2001) Mitogen-activated protein kinases are in vivo transducers of osmosensory signals in fish gill cells, Comp Biochem Physiol Part B Biochem Mol Biol 129, 821–829. 21 Lionetto MG & Schettino T (2006) The Na + -K + -2Cl cotransporter and the osmotic stress response in a model salt transport epithelium. Acta Physiol 187, 115–124. 22 Daborn K, Cozzi RRF & Marshall WS (2001) Dynam- ics of pavement cell–chloride cell interactions during abrupt salinity change in Fundulus heteroclitus. J Exp Biol 204, 1889–1899. 23 Lionetto MG, Pedersen SF, Hoffmann EK, Giordano ME & Schettino T (2002) Roles of the cytoskeleton and of protein phosphorylation events in the osmotic stress response in EEL intestinal epithelium. Cell Physiol Biochem 12, 163–178. 24 Marshall WS, Ossum CG & Hoffmann EK (2005) Hypotonic shock mediation by p38 MAPK, JNK, PKC, FAK, OSR1 and SPAK in osmosensing chloride secret- ing cells of killifish opercular epithelium. J Exp Biol 208, 1063–1077. 25 Ollivier H, Pichavant K, Puill-Stephan E, Roy S, Calves P, Nonnotte L & Nonnotte G (2006) Volume regulation D. F. Fiol and D. Ku ¨ ltz Osmotic stress sensing and signaling in fishes FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS 5797 following hyposmotic shock in isolated turbot (Scoph- thalmus maximus) hepatocytes. J Comp Physiol B Biochem Syst Environ Physiol 176, 393–403. 26 Fiol DF, Chan SY & Ku ¨ ltz D (2006) Identification and pathway analysis of immediate hyperosmotic stress responsive molecular mechanisms in tilapia (Oreochr- omis mossambicus) gill. Comp Biochem Physiol D Genomics Proteomics 1, 344–356. 27 Ku ¨ ltz D, Chakravarty D & Adilakshmi T (2001) A novel 14-3-3 gene is osmoregulated in gill epithelium of the euryhaline teleost Fundulus heteroclitus. J Exp Biol 204, 2975–2985. 28 Kohn A, Chakravarty D & Ku ¨ ltz D (2003) Teleost Fh14-3-3a protein protects Xenopus oocytes from hy- perosmolality. J Exp Zool Part A Comp Exp Biol 299A, 103–109. 29 Woo SK, Lee SD & Kwon HM (2002) TonEBP tran- scriptional activator in the cellular response to increased osmolality. Pflugers Arch 444, 579–585. 30 Lo ´ pez-Bojorquez LN, Villalobos P, Garcı ´ a GC, Orozco A, Valverde R & C (2007) Functional identification of an osmotic response element (ORE) in the promoter region of the killifish deiodinase 2 gene (FhDio2). J Exp Biol 210, 3126–3132. 31 Fiol DF & Ku ¨ ltz D (2005) Rapid hyperosmotic coin- duction of two tilapia (Oreochromis mossambicus) tran- scription factors in gill cells. Proc Natl Acad Sci USA 102, 927–932. 32 Fiol DF, Chan SY & Ku ¨ ltz D (2006) Regulation of osmotic stress transcription factor 1 (Ostf1) in tilapia (Oreochromis mossambicus) gill epithelium during salin- ity stress. J Exp Biol 209, 3257–3265. 33 Fiol DF, Mak SK & Ku ¨ ltz D (2006) Specific TSC22 domain transcripts are hypertonically induced and alternatively spliced to protect mouse kidney cells dur- ing osmotic stress. FEBS J 274, 109–124. 34 Ku ¨ ltz D (2005) Molecular and evolutionary basis of the cellular stress response. Annu Rev Physiol 67, 225–257. 35 Pan F, Zarate J & Bradley TM (2002) A homolog of the E3 ubiquitin ligase Rbx1 is induced during hyperos- motic stress of salmon. Am J Physiol Regul Integr Comp Physiol 282, R1643–R1653. 36 Ichimura T, Yamamura H, Sasamoto K, Tominaga Y, Taoka M, Kakiuchi K, Shinkawa T, Takahashi N, Shimada S & Isobe T (2005) 14-3-3 proteins modulate the expression of epithelial Na + channels by phosphory- lation-dependent interaction with Nedd4-2 ubiquitin ligase. J Biol Chem 280, 13187–13194. 37 Foskett JK, Bern HA, Machen TE & Conner M (1983) Chloride cells and the hormonal control of teleost fish osmoregulation. J Exp Biol 106, 255–281. 38 McCormick SD & Bradshaw D (2006) Hormonal con- trol of salt and water balance in vertebrates. Gen Comp Endocrinol 147, 3–8. 39 Evans DH, Piermarini PM & Choe KP (2005) The mul- tifunctional fish gill: dominant site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous waste. Physiol Rev 85, 97–177. 40 Evans DH, Rose RE, Roeser JM & Stidham JD (2004) NaCl transport across the opercular epithelium of Fundulus heteroclitus is inhibited by an endothelin to NO, superoxide, and prostanoid signaling axis. Am J Physiol Regul Integr Comp Physiol 286, R560–R568. 41 Hyndman KA, Choe KP, Havird JC, Rose RE, Pierma- rini PM & Evans DH (2006) Neuronal nitric oxide syn- thase in the gill of the killifish, Fundulus heteroclitus . Comp Biochem Physiol Part B Biochem Mol Biol 144, 510–519. Osmotic stress sensing and signaling in fishes D. F. Fiol and D. Ku ¨ ltz 5798 FEBS Journal 274 (2007) 5790–5798 ª 2007 The Authors Journal compilation ª 2007 FEBS . extracellular signaling. Overall, current insight into osmosensing and osmotic stress- induced signal transduction in fishes is limited. However, euryhaline fish. mitogen-activated protein kinase kinase kinase 7 interacting protein 2 (TAK 1 binding protein 2 ¼ TAB 2), as an IEG during hyperosmotic stress in tilapia gill epithelium