Dopaminergic regulation of ion transport in gills of the euryhaline semiterrestrial crab Chasmagnathus granulatus: interaction between D1- and D2-like receptors

Active ion absorption through the posterior gills of euryhaline crabs is regulated by bioamines and peptidic factors secreted by neuroendocrine organs such as pericardial organs (PO), thoracic ganglion (ThG) and sinus glands(Eckhardt et al., 1995; Kamemoto and Oyama, 1985; Mantel, 1985; Onken et al., 2000; Spanings-Pierrot et al., 2000;among others). In early works, PO extract, dopamine (DA), serotonin,octopamine and hemolymph from seawater acclimated crabs have been injected into whole animals or tested in isolated perfused gills of the blue crab Callinectes sapidus and the Chinese crab Eriocheir sinensis,resulting in increased Na⁺ influx and intracellular cyclic adenosine 3′-5′ monophospate (cAMP) content(Kamemoto, and Oyama, 1985; Lohrmann and Kamemoto, 1987; Sommer and Mantel, 1988; Sommer and Mantel, 1991; Bianchini and Gilles, 1990). The activity of gill Na⁺/K⁺-ATPase was also tested after injection of hemolymph from 30% seawater crabs(Savage and Robinson, 1983),injection of DA and PO extract (Sommer and Mantel, 1988), injection of DA or dibutyryl-cAMP(Morris and Edwards, 1995) and after incubation of gills with cAMP (Mo et al., 1998). All of these experiments have led to short-term activation of Na⁺/K⁺-ATPase, and it was concluded that DA activates Na⁺/K⁺-ATPase in a cAMP-dependent fashion,with the final result of increasing the rate of active Na⁺ uptake across the gills of euryhaline aquatic crabs experiencing hypo-osmotic stress(reviewed by Morris, 2001). Moreover, an increase in the DA content of hemolymph and gills of crabs transferred from saltwater to dilute medium has been shown(Zatta, 1987), suggesting that the DA-cAMP-Na⁺/K⁺-ATPase mechanism is relevant during in vivo, physiological conditions.

In addition, cAMP induces an increase in the short circuit current(I_sc) in split gill lamellae of the Chinese crab mounted in an Ussing chamber (Riestenpatt et al.,1994). However, Riestenpatt et al. concluded that the effect was not mediated via modulation of Na⁺/K⁺-ATPase activity, but that it was instead mostly due to an increase in apical conductivity for Na⁺ and Cl^-, together with an increase in the electromotive force for Cl^- entry. Although the reasons for the discrepancies regarding the Na⁺/K⁺-ATPase modulation by cAMP are not clear, this reveals that the regulation of ion transport by DA in crab gills is a complex mechanism that affects several ion-transporting proteins.

Chasmagnathus granulatus (Dana 1851) is a hyper-hypo-regulating crab inhabiting estuarine environments of Brazil, Uruguay and Argentina(Boschi, 1964; Charmantier et al., 2002). It is frequently exposed to dilute seawater as well as to rain pools, and is able to compensate the salt loss by active and electrogenic ion absorption through the posterior gills, with the Na⁺/K⁺-ATPase being the main driving force (Luquet et al.,2002; Onken et al.,2003). Recent work from our laboratory(Halperin et al., 2004)demonstrated that DA produces a transient increase of the transepithelial potential difference (V_te) in isolated perfused posterior gills of this species, with maximal effect dose between 10 and 50 μmol l^-1. This response is mediated by the cAMP-protein kinase A (PKA)pathway since it can be blocked by the PKA specific inhibitor KT5720, and is accompanied by a transient increase in Na⁺/K⁺-ATPase activity. These results and our preliminary data suggest that DA modulates ion transport activity in the gills of C. granulatus through a complex mechanism, which involves a stimulatory phase mediated by cAMP-PKA and a rapid deactivation phase that brings both V_te and Na⁺/K⁺-ATPase to resting values. These two regulatory phases could be explained by the interaction of two DA receptors linked to different G proteins, such as the D1 and D2 subtypes, extensively reported to modulate Na⁺/K⁺-ATPase activity in vertebrate tissues(for reviews, see Missale et al.,1998; Therien and Blostein,2000) and already pharmacologically detected in crustacean tissues(Kuo et al., 1995; Wilkens et al., 1996; Fingerman, 1997).

In vertebrates, D1-like receptors have been defined as those stimulating adenylyl cyclase upon stimulation with DA, while D2-like DA receptors are those that inhibit adenylyl cyclase(Missale et al., 1998). Recently, three groups of invertebrate DA receptors have been characterized,particularly in insects and in Caenorhabditis elegans. Two of these groups respond as D1-like, since they increase intracellular cAMP when stimulated with DA (Mustard et al.,2005). In particular, Trausch et al. have been able to block DA effects in the gills of the Chinese crab with the DA antagonist chlorpromazine(Trausch et al., 1989) and Morris and Edwards inhibited gill Na⁺/K⁺-ATPase activity in the amphibious crab Leptograpsus variegatus by injecting the DA antagonist butaclamol hydrochloride(Morris and Edwards, 1995). More recently, Mo et al. pharmacologically identified D1-like receptors in the gills of the E. sinensis (Mo et al., 2002). However, to our knowledge, little is known about the physiological effects mediated by the different subtypes of DA receptors in crab gills. The coupling of these receptors with intracellular signal transduction pathways also remains unexplored.

Our objective was to study the effects of DA on ion absorption (estimated from V_te) in isolated gills of C. granulatussubjected to different physiological conditions. To achieve a stimulated state of ion transport, gills were perfused with a hypo-osmotic solution, known to produce an increase in the V_te, I_scand Na⁺/K⁺-ATPase activity(Tresguerres et al., 2003), or with solutions containing cAMP agonists, which increase V_te, Na⁺ influx and Na⁺/K⁺-ATPase in isolated gills(Halperin et al., 2004). We also investigated the signal transduction pathways involved in the V_te response to DA, as well as the effects of agonists and antagonists of D1 and D2 dopamine receptor groups. Our results suggest that DA can either stimulate or inhibit ion uptake across C. granulatus gills depending on the gill physiological condition, and that at least part of this modulation takes place through an interaction between D1-like and D2-like receptors, the cAMP-PKA pathway and protein phosphatases.

Adult male crabs Chasmagnathus granulatus (Dana 1851) in intermolt stage C (Drach and Tchernigovtzeff,1967) were collected at Punta Rasa, San Clemente del Tuyú,Argentina. Animals were kept in plastic containers with aerated artificial brackish water of 10‰ salinity, fed twice a week with pellets of rabbit food and kept at room temperature of 20±1°C under a 12 h:12 h L:D photoperiod. Crabs of 30-33 mm carapace width were chosen for the study.

Crabs were sacrificed by destroying the nervous system with large scissors. After removing the dorsal carapace, gills number 6 (the largest among posterior gills) were gently excised and prepared according to Siebers et al.(Siebers et al., 1985). The afferent and efferent vessels were connected by fine polyethylene tubing(Ø=0.4 mm) to a peristaltic pump (afferent) and to a collecting tube(efferent). The tubing was held in position by an acrylic clamp covered with smooth neoprene to minimize gill damage. The preparation was bathed in a beaker containing 25 ml of aerated saline and was perfused at a rate of 0.1 ml min^-1. Under these conditions the preparations remained viable for several hours; gills not showing stable V_te after the usual stabilization period were discarded.

Ag/AgCl electrodes were connected via agar bridges to the external bath and to the collecting tube (internal side). Potential differences(outside-inside) were recorded using a millivoltmeter and a chart recorder.

Gill perfusions were performed in symmetrical conditions with isosmotic 30‰ saline solution (Table 1) until V_te was stabilized (about 30 min). This perfusion protocol led to a low V_te value, which was considered as reflecting the `steady state' conditions. Then, the drugs and drug combinations corresponding to each experiment were applied dissolved in the perfusion saline at the final concentrations detailed below. DMSO(0.05-0.1%) or ethanol (0.6%), which were used as vehicles for some drugs,were previously tested alone and did not affect V<sub>te</sub>. In some experiments, the perfusion saline was changed to a hypo-osmotic saline(20‰), which resembles the hemolymph composition of crabs acclimated to oligohaline medium (Charmantier et al.,2002; Tresguerres et al.,2003) to reach `stimulated' conditions before applying the specific treatment.

Dopamine (10 or 50 μmol l^-1) was a gift from C. J. Calvete,Fabra, Argentina. Okadaic acid (C₄₄H₆₇O₁₃Na,60 nmol l^-1, predissolved in ethanol) was a gift from Alomone Labs,Jerusalem, Israel; NaCl, KCl, CaCl₂, MgCl₂ and glucose were purchased from Merck (Buenos Aires, Argentina). NaHCO₃, and EDTA were from Mallinckrodt (New York, NY, USA). Hepes, forskolin (10 μmol l^-1, predissolved in DMSO), SCH23390 (10 μmol l^-1,predissolved in distilled water), fenoldopam (15 μmol l^-1,predissolved in distilled water), and spiperone (50 μmol l^-1,predissolved in DMSO) were obtained from Sigma (St Louis, MO, USA). Domperidone (10 μmol l^-1, predissolved in DMSO), ethanol and DMSO were purchased from ICN Biomedical Inc. (Irvine, CA, USA), Sintorgan(Buenos Aires, Argentina), and Carlo Erba (Milan, Italy) respectively. Theophylline (2.5 mmol l^-1) and Tris were from Serva (Heidelberg,Germany).

Data were analyzed by repeated-measures analysis of variance (RM-ANOVA),followed by Newman-Keuls multiple comparisons. All data are presented as mean± standard error of mean (s.e.m.). Differences were considered significant at P<0.05.

In gills perfused in steady state conditions (saline 30‰), the addition of 10 μmol l^-1 DA to the perfusate produced an initial increase of about 120% in V_te. This effect was spontaneously reversed and V_te returned to control values between 40 and 60 min after the start of the treatment(Fig. 1A,B). Similar effects were seen when gills were perfused with 50 μmol l^-1 DA (data not shown).

When the perfusion saline was changed to hypo-osmotic 20‰ saline, V_te was significantly hyper-polarized from 1.5±0.29 mV to 6.4±1.2mV, bath-positive (stimulated conditions). This effect lasted for as long as the hypo-osmotic saline was applied, in accordance with Tresguerres et al. (Tresguerres et al.,2003). Under these conditions, perfusion with 10 μmol l^-1 DA produced inconsistent effects on V_te. Increasing the DA concentration to 50 μmol l^-1 caused a reduction of 54% of the stimulation caused by the hypo-osmotic medium. Wash-out with fresh hypo-osmotic saline reversed the effect of DA(Fig. 2).

In a similar experiment, perfusion with the adenylyl cyclase activator forskolin pre-dissolved in DMSO (Laurenza et al., 1989) produced a marked increase of V_te from 2.0±0.23 mV to 12.89±1.84 mV. Addition of 50 μmol l^-1 DA to the perfusion saline reversed the stimulatory effect of forskolin by 25%(Fig. 3). Removing DA produced a recovery of the stimulated V_te.

Okadaic acid (OA) is a specific, potent and cell permeant inhibitor of PP1 and PP2A protein phosphatases (Cohen et al., 1990). Pre-treatment of perfused gills with 60 nmol l^-1 OA did not affect the steady state V_te. DA applied together with OA caused a significantly more pronounced and prolonged stimulation than DA alone. Furthermore, the presence of OA almost completely blocked the second (inhibitory) phase of the DA effect(Fig. 4).

This experiment was designed to test whether the inhibitory phase of DA can be prevented by maintaining high intracellular concentrations of cAMP. Theophylline has been used as an effective inhibitor of phosphodiesterase in crab gill preparations (Onken et al.,2000; Tresguerres et al.,2003). When this drug was applied to perfused gills of C. granulatus it produced a slight increase in V_te.Addition of DA in the presence of theophylline produced a further hyperpolarization, which was stable for at least 60 min and remained stable at the same level even when DA was removed. Applying DA after removing theophylline rapidly reduced V_te to control values,showing that the DA transduction system was intact(Fig. 5).

To test whether the stimulatory and inhibitory effects of DA were executed by binding of this bioamine to two different receptors, we applied agonists and antagonists of the DA receptor groups 1 and 2 (D1 and D2). The D1 antagonist SCH23390 (Iorio et al.,1983) was applied at a final concentration of 10 μmol l^-1 before and during perfusion with DA, under steady state conditions. The stimulatory phase was not affected while the inhibitory phase was almost totally blocked (Fig. 6A). Addition of 15 μmol l^-1 of the D1 agonist fenoldopam (Hussain and Lokhandwala,1997) reversed the stimulating effect of forskolin by 33%(Fig. 6B). The effect of fenoldopam on forskolin-stimulated gills was very similar to the effect previously obtained by perfusing DA in the same conditions(Fig. 3).

Finally, two D2 receptor antagonists, spiperone (50 μmol l^-1)and domperidone (10 μmol l^-1)(Kuo et al., 1995; Rodriguez et al., 2002), added before and during DA perfusion, completely blocked the DA effects. No DAdependent stimulation was seen after applying either of these antagonists. As a positive control, at the end of the experiments the drugs were washed out with fresh saline solution; then if the gills had been previously treated with domperidone, DA added alone produced the stimulatory effect described in previous experiments. On the other hand, the effect of spiperone was irreversible, since no further effect of DA could be observed after washing out the antagonist (Fig. 7A,B;refer to Fig. 1).

Dopaminergic regulation of gill ion transport has been studied in many euryhaline crab species with different respiratory habits. The general trend,with the exception of the fully terrestrial species, is that DA stimulates Na⁺/K⁺-ATPase and other ion-transporting proteins through an increase in the intracellular cAMP(Morris, 2001).

A recent paper of our laboratory(Halperin et al., 2004) shows that in C. granulatus, DA produces a transient increase in the transepithelial potential difference (V_te) and in the rate of ²²Na uptake in isolated gills of this species perfused with a 30‰ saline solution. Furthermore, our results suggest that the effects are mediated by a cAMP-PKA-dependent stimulation of Na⁺/K⁺-ATPase. We have also established that perfusion of C. granulatus gills with hypo-osmotic saline solution (20‰)strongly stimulates V_te and short-circuit current, at least in part through a cAMP-dependent mechanism that involves increased Na⁺/K⁺-ATPase activity(Tresguerres et al.,2003).

Since the response to both DA and hypo-osmotic perfusion seems to affect the same cellular components and to be mediated by similar cell signaling events, we studied the interaction between both stimuli. For this purpose we have defined `steady state conditions' when the gills are perfused with 30‰ saline solution, which is isosmotic to the hemolymph of estuarine and seawater acclimated crabs, and `stimulated conditions' when the gills are perfused with a hypo-osmotic saline (20‰), similar to the hemolymph of crabs acclimated to oligohaline medium. The steady state conditions are characterized by a low rate of adenylyl cyclase activity, since the addition of the phosphodiesterase inhibitors IBMX and theophylline produces much lower stimulation than perfusion with hypo-osmotic solution or with the adenylyl cyclase activator forskolin (Tresguerres et al., 2003; Halperin et al.,2004; present study).

DA applied to posterior gills of C. granulatus under steady state conditions causes a transient stimulation of V_te, as previously reported (Halperin et al.,2004). In contrast, upon stimulation by either hypo-osmotic perfusion or by the addition of forskolin, DA reduces V_te. This suggests that the inhibitory phase of DA effect occurs through an active component, which antagonizes the effects of other cAMP-mediated stimuli, and not merely through receptor saturation or deactivation. The blockade of this phase by the phosphodiesterase inhibitor theophylline (see Fig. 5) suggests that DA acts upstream of PKA, likely by inhibiting adenylyl cyclase. However, direct measurements of cAMP levels are needed to support this idea.

The role of protein phosphatases (PPs) in modulation of Na⁺/K⁺-ATPase has not yet been reported in crustacean gills. In contrast, there are many reports for other ion transporting epithelia like renal tubules of mammals. In these cases the PPs are known to dephosphorylate the Na⁺/K⁺-ATPase, regulating pump activity (Ewart and Klip,1995). Regulation of these enzymes by DA is also known to exist in mammalian kidneys e.g. through the regulatory protein DARP-32(Bertorello and Katz, 1995; Therien and Blostein, 2000). Our results support the involvement of PPs in the regulation of Na⁺/K⁺-ATPase by DA in C. granulatus gills,since their inhibition with okadaic acid blocks the inhibitory phase of the DA effect. Furthermore, since okadaic acid also reinforces the stimulating effect of DA, we can speculate that the PPs are already active under steady state conditions. In order to confirm this model, further experiments testing a possible direct relationship between DA on PPs activity are necessary.

There are many reports on dopaminergic regulation of Na⁺/K⁺-ATPase activity mediated by protein kinases in different vertebrate tissues. Phosphorylation of Na⁺/K⁺-ATPase on residues specific for PKA or protein kinase C (PKC) stimulates or inhibits the enzyme's activity in a tissue specific fashion (Therien and Blostein,2000). Although no study has proven direct phosphorylation of crab gill Na⁺/K⁺-ATPase, two reports give indirect support to this possibility. Trausch et al. have reported that cAMP, DA and serotonin induce the phosphorylation of proteins collected in the cell fraction expected to contain most of the Na⁺/K⁺-ATPase activity(Trausch et al., 1989). In the same study, the induced phosphorylation was correlated to increased Na⁺/K⁺-ATPase activity in db-cAMP perfused gills of E. sinensis. In addition, Towle et al. identified one putative site for PKA and several putative sites for PKC phosphorylation in the sequence of Callinectes sapidus Na⁺/K⁺-ATPaseα-subunit (Towle et al.,2001). These phosphorylation sites are probably involved in modulating the enzyme activity and/or its targeting to the cell membrane.

Using several of the D1 and D2 effectors most commonly used in vertebrate studies, we have been able to mimic or block both the stimulating and the inhibiting effects of DA, thus supporting the existence of different DA receptor subtypes in the gills of C. granulatus. However, the effects of these drugs are clearly different from the pattern described in the literature (Bertorello and Katz,1995; Missale et al.,1998; Therien and Blostein,2000; Asghar et al.,2001; Mustard et al.,2005). In the gills of C. granulatus, domperidone and spiperone, which are known D2 antagonists in vertebrates, seem to function as D1-like antagonists, since they produce the same effect as the one produced by the PKA inhibitor KT5720, a total blockade of the DA-induced stimulation of V_te. Accordingly, Trausch et al. have inhibited the DA induced, PKA-dependent phosphorylation of proteins by adding chlorpromazine another vertebrate D2 receptor antagonist(Trausch et al., 1989). However, these authors have not discussed their results in terms of receptor subtypes. It must be taken into account that in the few DA receptors characterized in invertebrates, particularly in insects, spiperone is effective to inhibit the D1-like response at μmolar concentrations, as in the results of the present paper, whereas the effectiveness of this drug has been tested in only one D2-like invertebrate DA receptor, with weaker effects(Mustard et al., 2005). On the other hand, domperidone is also able to block the D1-like response in C.granulatus but is not effective on the insect D1-like receptor AmDop2 (Mustard et al.,2005).

SCH23390, a D1-like antagonist both in vertebrates and insects(Missale et al., 1998; Mustard et al., 2005) blocks most of the inhibitory (D2-like) phase of the DA effect on steady state gills of C. granulatus, in the same way as the inhibitors of phosphodiesterase and PPs, theophylline and OA. In the same experiments,SCH23390 does not affect the stimulatory (D1-like) phase. These results are partially coincident with the results obtained for C. elegans, in which SCH23390 acts as an antagonist of the D2-like receptor CeDOP2 and as agonist of the D1-like CeDOP1 (for a review, see Mustard et al., 2005). The D1 agonist fenoldopam mimics the inhibitory phase of DA effect, reducing V_te on gills of C. granulatus previously stimulated with forskolin, in contrast to the report(Hussain et al., 1997) in which bromocriptine (D2 agonist) partially reverts the effect of forskolin on the Na⁺/K⁺-ATPase of rat renal proximal tubules. D1-like receptors have already been identified in the gills of E. sinensis by studying the binding of the D1 antagonist SCH23390(Mo et al., 2002). However,our results suggest that in crabs SCH23390 could be a D2-like and not a D1-like antagonist. Table 2summarizes the effects of several DA effectors on vertebrates and invertebrates. Comparative pharmacology of DA receptors is difficult to interpret since the response of the few characterized invertebrate (insect)receptors matches with that in vertebrates only for some effectors like the D1 antagonist SCH23390. By contrast spiperone, which is an effective D2 antagonist in vertebrates, antagonizes D1-like receptors more efficiently than D2-like receptors in invertebrates(Missale et al., 1998; Mustard et al., 2005). As far as we can conclude from our results, the response of C. granulatus DA receptors is different from that in vertebrates for these two antagonists and also for domperidone and fenoldopam. Compared to other invertebrates, C. granulatus D1-like receptors respond to spiperone in the same way as in insects, while D2-like receptors respond to SCH23390 in a similar fashion as in C. elegans.

We propose that, in C. granulatus gills, DA stimulates adenylyl cyclase and therefore ion transport through D1-like receptors linked to activation of adenylyl cyclase. D2-like receptors appear to inhibit ion transport activity, estimated as V_te, through reduction of cAMP levels.

The fact that V_te is first increased and then decreased suggests that in the gills of C. granulatus, D1-like receptors have higher affinity than D2-like receptors, or the latter are activated by elevated cAMP levels. The first hypothesis offers an explanation for previous results (Halperin et al.,2004) in which DA concentrations higher than 50 μmol l^-1 produce less pronounced and less sustained activation of V_te in isolated perfused gills of this species. Perfusion with a high concentration of DA can rapidly activate D2-like receptors,antagonizing the response to the more sensitive D1-like receptors. This could also explain our results, in which the most stimulating DA concentration reported by Halperin et al. (10 μmol l^-1)(Halperin et al., 2004) is not enough to produce a consistent inhibitory effect on hypo-osmotically or forskolin stimulated gills, while 50 μmol l^-1 DA effectively reduces the hypo-osmotic- or forskolin-induced V_te. However, cAMP-mediated phosphorylation-deactivation of D1-like receptors cannot be ruled out from these hypotheses.

As for the role of DA in osmoregulation, it is clear that this neurohormone modulates the gill V_te in different ways, depending on the level of adenylyl cyclase activity. When this enzyme works at the steady state level, DA stimulates cAMP production, causing a transient increase in V_te, Na⁺/K⁺-ATPase activity and sodium transport. On the other hand, when gill adenylyl cyclase has already been activated by either DA, hypo-osmotic medium or any other stimulus involving cAMP transduction, the effect of DA is to reduce its activity level. Further work is required to characterize the molecular nature of dopamine receptors in crustacean gills and their role in osmoregulation.

List of abbreviations
 
• cAMP

  cyclic adenosine 3′-5′ monophospate

 
• D1 and D2

  DA receptor groups 1 and 2

 
• DA

  dopamine

 
• I_sc

  short circuit current

 
• OA

  okadaic acid

 
• PKA

  protein kinase A

 
• PKC

  protein kinase C

 
• PO

  pericardial organs

 
• PPs

  protein phosphatases

 
• s.e.m.

  standard error of mean

 
• ThG

  thoracic ganglion

 
• V_te

  transepithelial potential difference

This work was supported by CONICET grant PIP-02348 to CML and by National Science Foundation grant INT-0305484 to D.W.T. Martín Tresguerres holds an Isaac Walton Killam Memorial Scholarship at the Department of Biological Sciences, University of Alberta, Canada.