Hydrogen sulfide stimulates CFTR in Xenopus oocytes by activation of the cAMP/PKA signalling axis

Hydrogen sulfide (H2S) has been recognized as a signalling molecule which affects the activity of ion channels and transporters in epithelial cells. The cystic fibrosis transmembrane conductance regulator (CFTR) is an epithelial anion channel and a key regulator of electrolyte and fluid homeostasis. In this study, we investigated the regulation of CFTR by H2S. Human CFTR was heterologously expressed in Xenopus oocytes and its activity was electrophysiologically measured by microelectrode recordings. The H2S-forming sulphur salt Na2S as well as the slow-releasing H2S-liberating compound GYY4137 increased transmembrane currents of CFTR-expressing oocytes. Na2S had no effect on native, non-injected oocytes. The effect of Na2S was blocked by the CFTR inhibitor CFTR_inh172, the adenylyl cyclase inhibitor MDL 12330A, and the protein kinase A antagonist cAMPS-Rp. Na2S potentiated CFTR stimulation by forskolin, but not that by IBMX. Na2S enhanced CFTR stimulation by membrane-permeable 8Br-cAMP under inhibition of adenylyl cyclase-mediated cAMP production by MDL 12330A. These data indicate that H2S activates CFTR in Xenopus oocytes by inhibiting phosphodiesterase activity and subsequent stimulation of CFTR by cAMP-dependent protein kinase A. In epithelia, an increased CFTR activity may correspond to a pro-secretory response to H2S which may be endogenously produced by the epithelium or H2S-generating microflora.

The Na 2 S-induced current of CFTR-expressing oocytes was rapidly inhibited by the additional application of the CFTR inhibitor CFTR_inh-172 (Fig. 1d,e). Under perfusion with 50 µM Na 2 S, transmembrane currents increased to −0.757 ± 0.145 µA and were rapidly inhibited to −0.153 ± 0.039 µA after addition of 25 µM CFTR_ inh-172 (n = 8; N = 2; P = 0.0078). The transient nature of the current which was stimulated by Na 2 S was not the result of a rapid evaporative loss of H 2 S from the buffer solutions. Na 2 S was measured in the employed buffer solution by the formation of methylene blue and detection of its absorbance at 670 nm (Fig. 1f). There was only a minor decrease in methylene blue concentrations over time, indicating that H 2 S was present in the buffers even after a time period where current signals began to decline.
In order to confirm that the observed activation of CFTR by Na 2 S was due to H 2 S, the H 2 S-releasing compound GYY4137 which is chemically different from a sulphur salt 22 was employed. Since GYY4137 is a slow-releasing H 2 S donor 22 , higher concentrations (500 µM) were used (Fig. 1g,h). GYY4137 elicited a small but significant activation of CFTR. Transmembrane currents of CFTR expressing oocytes significantly increased from −0.046 ± 0.010 µA to −0.119 ± 0.020 µA (n = 7; N = 2; P = 0.0156) due to application of GYY4137 (Fig. 1g,h).
In sum, these data indicate that H 2 S activates human CFTR which is heterologously expressed in Xenopus oocytes.
Hydrogen sulfide stimulates CFTR activity via cAMP-mediated signalling events. The classical intracellular signalling cascade activating CFTR involves adenylyl cyclase (AC)-mediated production of cAMP and subsequent activation of protein kinase A (PKA). PKA phosphorylates the regulatory domain of CFTR and activates the channel in the presence of ATP. In order to investigate whether or not H 2 S interferes with this signalling pathway, the effect of the cAMP-elevating drugs forskolin/IBMX was evaluated with or without H 2 S. Depicted are values of I M (before drug application or peak values after drug application) from individual experiments (grey symbols) as well as means ± SEM (*P ≤ 0.05, Wilcoxon signed rank test; **P ≤ 0.01, Student's paired t-test). (c) Summarised data from experiments as similar to those shown in panels a and b, using native, non-CFTR-expressing oocytes. Values of I M were taken at time point were CFTR-expressing oocytes of the same donor had the maximal response to the drugs. Depicted are means ± SEM (**P ≤ 0.01, Student's paired t-test). (d) A representative current trace of a TEVC recording of a CFTR-expressing oocyte. After application of Na 2 S (50 µM, black bar), the CFTR inhibitor CFTR_inh172 (CFTR_ inh.; 25 µM) was additionally applied. This readily inhibited values of I M . (e) Statistical analysis of data obtained from experiments as shown in panel d. Depicted are values of I M (before drug application or peak values after drug application) from individual experiments (grey symbols) as well as means ± SEM (**P ≤ 0.01, Wilcoxon signed rank test). (f) Evaporative loss of H 2 S was measured by monitoring the concentration of H 2 S in the employed buffers solutions by the formation of methylene blue. Depicted are values for methylene blue absorbance at 670 nm over time. Na 2 S (50 µM) exposure is indicated by the black bar. (g) A representative current trace of a TEVC recording of a CFTR-expressing oocyte. Both, GYY4137 (500 µM, grey bar) as well as Na 2 S (50 µM, black bar) stimulated I M . (h) Statistical analysis of data obtained from experiments as shown in panel g. Depicted are values of I M (peak values after drug application) from individual experiments (grey symbols) as well as means ± SEM (*P ≤ 0.05, Wilcoxon signed rank test). Numbers of experiments (n) are indicated in parentheses.
The application of forskolin/IBMX elicited a significant and transient increase in transmembrane current (Fig. 2a) from −0.064 ± 0.009 µA to −1.364 ± 0.284 µA (n = 11; N = 5; P = 0.0009). This effect was fully reversible upon removal of the drugs. A second application of forskolin/IBMX again stimulated CFTR activity from −0.081 ± 0.022 µA to −0.604 ± 0.114 A (n = 11; N = 5; P = 0.0038; Fig. 2a). The second effect of forskolin/IBMX was normalised to the effect of the first forskolin/IBMX application and defined as 'normalised CFTR activity' . Without any additional treatment, control oocytes had thus a normalised CFTR activity of 0.42 ± 0.04 (n = 11; N = 5). We then applied increasing concentrations of Na 2 S after the first application of forskolin/IBMX (Fig. 2a). Interestingly, 50 µM Na 2 S which elicited robust currents in previous experiments ( Fig. 1) did not significantly stimulate transmembrane currents after the oocytes had been exposed to forskolin/IBMX (Fig. 2a,c). Only a high dose of 300 µM Na 2 S triggered a small increase in transmembrane currents from −0.033 ± 0.010 µA to −0.051 ± 0.009 µA (n = 8; N = 3; P = 0.0423; Fig. 2a,c). However, despite the lack of an effect of Na 2 S after previous exposure of the oocytes to forskolin/IBMX, Na 2 S enhanced the second effect of forskolin/IBMX. Normalised CFTR activity dose-dependently increased due to application of Na 2 S (Fig. 2a,d). This effect was inhibited by 25 µM of CFTR_inh.172 (Fig. 2b,d). Furthermore, there was only a minor current activation due to 50 µM Na 2 S and forskolin/IBMX in native, non-CFTR expressing oocytes (Fig. 1c). These data suggest that Na 2 S potentiates CFTR-activity which was elicited by forskolin/IBMX.
To investigate if the Na 2 S-induced stimulation of CFTR involves AC and PKA, specific inhibitors of these enzymes were employed and Na 2 S-induced currents (I Na2S ) were estimated with or without these drugs (Fig. 3). Na 2 S (50 µM) was applied twice to CFTR-expressing oocyte in order to control for a potential desensitisation in response to repetitive Na 2 S-exposure (Fig. 3a,b). The first I Na2S was 0.321 ± 0.108 µA and not significantly different from the second I Na2S which was 0.312 ± 0.087 µA ( Fig. 3b; n = 6; N = 5; P = 0.854).
A similar experiment was performed with cAMPS-Rp -a PKA antagonist -which was directly injected into the oocytes during experiments. For control experiments, CFTR-expressing oocytes were stimulated with Na 2 S (50 µM). Subsequently, 9.2 nl of an intracellular-analogous solution (IAS) was injected into the oocytes and the cells were stimulated a second time with Na 2 S. This manoeuvre increased (although values did not reach statistical significance) I Na2S from 0.255 ± 0.046 µA to 0.484 ± 0.126 µA (n = 9; N = 2; P = 0.0743; Fig. 3e,f). A similar observation has been reported in a previous study 23 , where a volume-increase in oocytes increased the activation of CFTR by cAMP-elevating compounds. By contrast, injection of the PKA antagonist cAMPS-Rp abrogated the second effect of Na 2 S. Values of I Na2S significantly decreased from 0.248 ± 0.061 µA to 0.011 ± 0.006 µA (n = 9; N = 2; P = 0.0039; Fig. 3e,f). Taken together, these data show that H 2 S activates CFTR by cAMP-and PKA-mediated signalling in Xenopus oocytes.
Hydrogen sulfide targets phosphodiesterase rather than adenylyl cyclase. An increase in intracellular cAMP concentrations could either be the result of enhanced cAMP production by AC, or inhibition of cAMP degradation by phosphodiesterase (PDE). H 2 S might thus stimulate AC or inhibit PDE -both effects would result in accumulation of cAMP, a downstream activation of PKA and subsequent stimulation of CFTR. In order to discriminate between AC-and PDE-mediated contributions to CFTR activation, CFTR was stimulated with maximal effective concentrations of either forskolin (AC activator) or IBMX (PDE inhibitor). If H 2 S potentiated the effect of forskolin, but not that of IBMX, H 2 S likely prevents cAMP degradation by PDE. If H 2 S potentiated the effect of IBMX, but not that of forskolin, H 2 S likely stimulates cAMP production by AC.
First, we investigated if H 2 S affects the effect of forskolin alone (Fig. 4a). Since we were not able to additionally stimulate CFTR activity by increasing the forskolin concentration to 30 µM (data not shown), we considered the employed concentration of 5 µM as maximally effective, an observation which is consistent with a reported EC 50 value of ~0.07 µM for forskolin in airway epithelia 24 . In control experiments, forskolin (5 µM) was applied to CFTR-expressing oocytes, which stimulated I M by 0.324 ± 0.056 µA (n = 13; N = 3). After wash-out, the oocytes were stimulated again with 5 µM forskolin. This resulted in a second stimulation of I M by 0.363 ± 0.068 µA (n = 13; N = 3; Fig. 4a). The second effect of forskolin was normalised to the effect of the first forskolin application and defined as 'normalised forskolin effect' (Fig. 4b). Under these control conditions, the normalised forskolin effect was 1.135 ± 0.146 (n = 13; N = 2). By contrast, oocytes which were treated with 50 µM Na 2 S (together with the second application of forskolin) had a significantly enhanced normalised forskolin effect of 3.054 ± 0.405 (n = 13, N = 3, P < 0.0001, Gaussian approximation; Fig. 4a,b).
An identical protocol was employed with a high concentration of the PDE inhibitor IBMX (1 mM) and a 'normalised IBMX effect' was estimated (Fig. 4c,d). The normalised IBMX effect was 0.714 ± 0.080 (n = 10; N = 3) under control conditions, and not significantly different from that which was estimated in the presence of Na 2 S which was 0.966 ± 0.178 (n = 10; N = 3; P = 0.2208). Na 2 S thus enhanced the effect of the AC-agonist forskolin, but not that of the PDE-inhibitor IBMX. H 2 S might therefore impair PDE-mediated cAMP degradation rather than AC-mediated cAMP production.
In order to confirm these observations a different strategy was employed ( Fig. 4e-g). AC-mediated cAMP-production was blocked by application of the AC inhibitor MDL 12330 A (20 µM) to CFTR-expressing oocytes. Subsequently, 100 µM of membrane-permeable 8-Br-cAMP was applied. This stimulated an increase in transmembrane current (I cAMP ) of 0.056 ± 0.0221 µA (n = 7; N = 4). After washout of all drugs, MDL 12330 A (20 µM) was applied again and 50 µM Na 2 S was added. Afterwards, 8Br-cAMP was additionally applied and I cAMP significantly increased to 0.590 ± 0.154 µA (n = 7; N = 4; P = 0.0124; Fig. 4e,g). By contrast, there was no difference between the first and second I cAMP (0.091 ± 0.026 µA and 0.064 ± 0.016 µA; n = 5; N = 2; P = 0.1561) when the procedure was repeated without Na 2 S (Fig. 4f,g).These data indicate that H 2 S enhances the efficacy of 8-Br-cAMP.
In sum, these data provide evidence that H 2 S inhibits endogenous PDE in Xenopus oocytes. This results in cAMP-mediated stimulation of CFTR-activity via downstream activation by PKA.

Discussion
In this study we investigated the regulation of CFTR by H 2 S. Previous studies using the mouse hippocampal cell line HT22 25 or rat vaginal epithelial preparations 21 suggested that CFTR might be a target for H 2 S. In order to elaborate on this hypothesis, human CFTR was heterologously expressed in Xenopus oocytes and functional CFTR expression was confirmed by application of the cAMP-elevating compounds forskolin and IBMX, which resulted in a transient increase in transmembrane currents which did not occur in native, non-injected oocytes. These observations are consistent with previously published functional electrophysiological analyses of human CFTR in Xenopus oocytes 23,26 . The H 2 S-liberating sulphur salt Na 2 S elicited a transient current stimulation of CFTR-expressing oocytes which was readily inhibited by the CFTR inhibitor CFTR_inh172 and did not occur in native oocytes. Furthermore, the slow-releasing H 2 S-liberating molecule GYY4137 also stimulated transmembrane currents in Xenopus oocytes. These data indicate that H 2 S, released from Na 2 S or GYY4137, stimulates CFTR activity.
We then elaborated on the signalling mechanisms which mediate the H 2 S-induced activation of CFTR. We first stimulated CFTR-expressing oocytes with forskolin/IBMX and after removal of these drugs, cells were exposed to H 2 S. Interestingly, a direct response to H 2 S only occurred when high concentrations of Na 2 S were employed. When forskolin/IBMX were applied twice to CFTR-expressing oocytes, the second activation by forskolin/IBMX was ~60%   (1) and second (2) Na 2 S-induced current (I Na2S ) from individual experiments (grey symbols) as well as means ± SEM (n.s. = not significant, Student's paired t-test). I Na2S was calculated by subtracting the current before application of Na 2 S from the peak value after application of Na 2 S, resulting in positive values for I NA2S . (c) Representative current traces of TEVC recordings of CFTRexpressing oocytes. Transmembrane currents (I M ) were recorded and oocytes were exposed twice to Na 2 S (50 µM, black bar) DMSO (0.1%; left trace) or the AC inhibitor MDL 12330 A (MDL, 20 µM; right trace) were applied between the first and second stimulation with Na 2 S (black arrowheads). (d) Statistical analysis of data obtained from experiments as shown in panel a. Depicted are values of the first (1) and second (2) Na 2 S-induced current (I Na2S ) from individual experiments (grey symbols) as well as means ± SEM (**P ≤ 0.01, Student's paired t-test). I Na2S was calculated by subtracting the current before application of Na 2 S from the peak value after application of Na 2 S, resulting in positive values for I NA2S . (e) Representative current traces of TEVC recordings of CFTR-expressing oocytes. Transmembrane currents (I M ) were recorded and oocytes were exposed twice to Na 2 S (50 µM, black bar). The perfusion recording was stopped briefly between the first and second stimulation with Na 2 S (grey lines). Then, an intracellular-analogous solution (IAS) or IAS containing the PK-antagonist cAMPS-Rp (87 µM) were injected into the oocytes before the second stimulation with Na 2 S (black arrowheads).  smaller than the first one. This is consistent with previous reports 23 and can be explained by a desensitisation in response to compounds which lead to a strong increase in intracellular cAMP 26 . The lack of a pronounced effect of H 2 S after initial stimulation with forskolin/IBMX is likely explained by the fact that Na 2 S alone is a much weaker stimulator of CFTR activity than forskolin/IBMX (as shown in Fig. 1a) and its effect is lost within the desensitisation after exposure to these compounds. Nevertheless, we found that H 2 S potentiated the second effect of forskolin and IBMX, indicating that the effect of H 2 S is additive to that of forskolin/IBMX. Furthermore, the H 2 S-induced activation of CFTR was lost in the presence of the AC inhibitor MDL12330A as well as the PKA antagonist cAMPS-Rp. These data indicate that H 2 S activates CFTR via the cAMP/PKA signalling axis, a finding which is consistent with a previous study demonstrating an increase in intracellular cAMP concentrations by H 2 S in Xenopus oocytes 27 . H 2 S is membrane-permeable 28 and might target AC in order to stimulate cAMP production. It has been previously shown that H 2 S either stimulates 27,29 or inhibits 30,31 AC activity, indicating that this enzyme represents indeed a molecular target for H 2 S. However, we found that H 2 S was able to potentiate the stimulation of CFTR by the AC-activator forskolin, but not by a high concentration (1 mM) of the PDE inhibitor IBMX. If H 2 S was acting on AC, there should be an additional effect over that of PDE inhibition alone -irrespective of the concentration of the PDE inhibitor. These data therefore suggest that H 2 S targets degradation rather than production of cAMP. However, these experiments alone do not justify a definite conclusion on the target for H 2 S. Based on the experiment using forskolin alone, a potential effect of H 2 S on AC cannot be ruled out completely. Forskolin increases the affinitiy of two cytoplasmic domains (C1 and C2) of AC for each other and enhances its activity 32 . In the presence of forskolin, AC becomes more sensitive to e.g. G sα 32 , suggesting that it is possible that there is a synergistic activation of AC by H 2 S in the presence of forskolin, but not in the presence of IBMX.
In addition to inhibiting PDEs, IBMX inhibits adenosine receptors, including endogenous adenosine receptors in Xenopus oocytes. Activation of these receptors inhibits AC and antagonising the receptor with IBMX might therefore elevate cAMP-concentrations -irrespective of the block of PDE-mediated degradation of cAMP. However, as shown in a study by Kobayashi et al. the receptors need to be activated by a ligand (adenosine) in order to generate a current signal in oocytes and this does not occur with a perfusion system as employed in our study 33 . Furthermore, it is not possible to discriminate between cGMP-and cAMP-PDEs using IBMX. It is possible that cGMP activates CFTR in Xenopus oocytes 34 , however, if H 2 S acted via cGMP, the effects of Na 2 S should not be sensitive to the adenylyl cyclase inhibitor MDL. This is further confirmed by the fact that H 2 S potentiated the stimulation of CFTR by exogenous membrane-permeable 8-Br-cAMP under conditions where endogenous cAMP production was blocked by the AC inhibitor MDL 12330A. The increased CFTR stimulation by 8-Br-cAMP in the presence of H 2 S can only be explained by impaired degradation of 8-Br-cAMP by endogenous PDE activity. Consistent with this hypothesis, several studies demonstrated that H 2 S is able to inhibit cAMP and cGMP degradation by PDEs [35][36][37] .
Most interestingly, small amounts (nano-to lower micro-molar range) of the sulphur salt NaHS inhibited PDE-mediated cAMP breakdown in cell-free systems 35,37 , suggesting that H 2 S might directly interfere with PDE activity -independently of cellular signalling cascades. This might be the result of an interference of H 2 S with zinc, thereby reducing the activity of zinc-dependent PDE 35 . Alternatively, H 2 S might target cysteine residues in the enzyme by persulfidation 35 . However, H 2 S alone cannot modify thiol groups 38 , whereas derivatives of H 2 S such as polysulfides 13 or nitroxyl, which might be formed in the cytoplasm of cells, are able to do so 38 . The precise mechanisms how H 2 S regulates PDE function remain to be elucidated; however, the Xenopus oocyte might be a useful model in order to address these questions.
In sum, we provide evidence that H 2 S inhibits endogenous PDE in the Xenopus oocyte, which likely results in an accumulation of cAMP and downstream activation of CFTR via PKA. This scenario, however, requires a constitutive production of cAMP in these cells. Xenopus oocytes are arrested in the G2 stage of meiosis I 39 . The G2 arrest is maintained by AC-mediated production of cAMP which is believed to prevent oocyte maturation via PKA-mediated signalling events [39][40][41] . Hence, there is a constitutive production of cAMP in Xenopus oocytes. Furthermore, there is an endogenous PDE activity in these cells 42 and the activities of both AC as well as PDE determine the concentration of cAMP. When Xenopus oocytes are exposed to progesterone, cAMP levels decrease, the cells undergo nuclear membrane breakdown (GVBD) and mature into a fertilisable egg 39 . This process is inhibited by IBMX 42 , again demonstrating that inhibition of PDE activity leads to an increase in the concentration of cAMP. Consistently, IBMX as well as H 2 S alone were able to stimulate CFTR activity in these cells in the present study.
Our data are consistent with an emerging body of evidence that H 2 S targets the cAMP-signalling axis. Whereas this study and others 27,29 provide evidence for an activation of the cAMP-pathway by H 2 S, inhibition of cAMP-signalling has also been reported [43][44][45] . A study by Lu et al. demonstrated inhibition of AC and stimulation of PDE by NaHS in AS4.1 cells 44 . These inconsistent findings suggest that the net-effect of H 2 S on cAMP signalling depends not only on whether AC or PDE is targeted by H 2 S, but also on the precise molecular repertoire of the cAMP-regulating machinery. In humans, ten isoforms of AC have been identified and there are 11 members of the PDE protein family which can generate nearly 100 different subtypes 46 . The specific isoform expression in a cell might critically determine whether H 2 S activates or inhibits cAMP-signalling. This is especially important, since H 2 S will diffuse across cell membranes in an unspecific manner and hence specificity is likely not achieved by membrane receptors but possibly by the subtype repertoire of intracellular signalling molecules.
In epithelia, H 2 S exerts pro-secretory or anti-absorptive effects 15,47 and we recently suggested a concept by which epithelia use their electrolyte and liquid transport machinery as a defence mechanism in order to flush potential sources for harmful amounts of H 2 S from the epithelial surfaces 15 . The herein reported cAMP-mediated stimulation of CFTR activity would be consistent with a pro-secretory action on chloride-secreting epithelia. A recent study demonstrated a CFTR-mediated chloride secretion across rat vaginal epithelial preparations 21 . The authors speculated that this might be due to an increase in cAMP 21 and our findings would support this hypothesis. Since H 2 S is not directly targeting CFTR, the PDE repertoire of epithelial cells will determine whether or not H 2 S triggers CFTR-mediated electrolyte secretion.
SCIeNTIfIC RepoRts | 7: 3517 | DOI:10.1038/s41598-017-03742-5 Aside from the pro-secretory effects, H 2 S prevents liquid absorption by sodium-transporting epithelia [48][49][50] . In these epithelia, cAMP/PKA signalling stimulates sodium absorption by an increase in the membrane abundance of sodium transporting molecules such as the epithelial sodium channel (ENaC) 51,52 . Interestingly, we recently showed that H 2 S prevents this cAMP-mediated increase in ENaC abundance in sodium-absorbing lung epithelial cells by yet unidentified mechanisms 50 . Furthermore, H 2 S did not enhance cAMP-mediated chloride secretion in primarily sodium absorbing pig airway surface epithelia (data not shown). Nevertheless, this illustrates that -depending on the specific enzymatic repertoire of the cAMP signalling axis -H 2 S might trigger cAMP-mediated electrolyte secretion in a fraction of epithelial cells, whilst simultaneously preventing enhanced electrolyte absorption in other cells. The herein reported data thus provide a step further in understanding the mechanisms of how H 2 S elicits a switch from absorptive to secretory electrolyte and water transport in epithelia.

Methods
Isolation of Xenopus laevis oocytes. All animal experiments were performed in accordance with the German animal welfare law and had been declared to the Animal Welfare Officer of the University (Registration No.: M_ 478 and M_549). The animal housing facility was licensed by the local authorities (Az: FD 62 - §11 JLU Tierphysiologie). The methods used to euthanize the animals humanely were consistent with the recommendations of the AVMA Guidelines for the Euthanasia of Animals. All procedures and experimental protocols were approved by the Animal Welfare Officer of the University as well as the regional council of Giessen (Registration No.: M_ 478 and M_549).
Oocytes of stages V/VI were isolated from freshly dissected Xenopus laevis ovaries and defolliculated exactly as previously described 53

Microelectrode recordings (Two-Electrode Voltage-Clamp, TEVC). CFTR-expressing oocytes
were placed in a Lucite chamber which was continuously perfused with oocyte ringer solution (ORS) containing 90 mM NaCl, 1 mM KCl, 2 mM CaCl 2 , 5 mM HEPES at pH 7.4. Chlorinated silver wires served as recording electrodes and were mounted into borosilicate glass capillaries (Hilgenberg, Malsfeld, Germany) with an outer diameter of 1.2 mm, which were pulled to microelectrodes with a DMZ universal puller (Zeitz-Instruments, Martinsried, Germany) and filled with 1 M KCl. Ag/AgCl wires were used as reference electrodes and placed directly into the recording chamber. The membrane voltage was clamped to −60 mV using a TEVC amplifier (Warner Instruments, Hamden, Connecticut, USA). Transmembrane currents (I M ) were low-pass filtered at 1000 Hertz (Frequency Devices 902, Haverhill, Massachusetts, USA) and continuously recorded with a strip chart recorder (Kipp&Zonen, Delft, The Netherlands).
Determination of evaporative loss of H 2 S from buffer solutions. The equilibration of H 2 S with its concentration in air will eventually lead to evaporative loss of this gas from the experimental buffer solutions 54 . Therefore the relative concentration of H 2 S in ORS was determined at various time points after administration of 50 µM Na 2 S by the formation of methylene blue. Samples (300 µl) of the solutions were mixed with 500 µl of 4% zinc acetate and incubated on ice for at least 30 min. Afterwards, 200 µl of 0.1% dimethylphenylendiamine sulfate (in 5 M HCl) and 100 µl of 50 mM FeCl 3 were added. Samples were vortexed, centrifuged at 5000× g and incubated for 5 min at room temperature. The absorption of methylene blue was measured at 670 nm with a Vis-spectrophotometer (Kruess Optronic, Hamburg, Germany).

Chemicals and solutions.
In order to apply H 2 S, the sulfur salt Na 2 S (Sigma, Taufkirchen, Germany) or the slow releasing H 2 S donor GYY4137 (Santa Cruz, Biotechnology, Dallas, Texas, USA) were employed. Na 2 S was prepared as a stock solution of 50 mM in ORS freshly before experiments and immediately diluted to final working concentrations in order to prevent evaporative loss of H 2 S from the experimental solutions. Stock solutions of 100 mM GYY4137 were prepared in H 2 O and stored at −20 °C. Forskolin (MoBiTec, Göttingen, Germany) was used as a stimulator of adenylyl cyclase and stock solutions of 10 mM were prepared in dimethyl sulfoxide (DMSO, Sigma) and stored at −20 °C. The phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX; Sigma, Taufkirchen, Germany) was dissolved to 100 mM in DMSO and stored at +4 °C. The adenylyl cyclase inhibitor MDL 12330 A hydrochloride (MDL; Tocris Bioscience, Bristol, UK) was dissolved to 20 mM in DMSO and stored at +4 °C. cAMPS-Rp (Tocris Bioscience) was used as a competitive antagonist of cAMP-induced Protein Kinase A (PKA). Stock solutions of cAMPS-Rp were prepared to 10 mM in H 2 O and stored at −20 °C. cAMPS-Rp was injected into Xenopus oocytes during TEVC experiments. Therefore, stock solutions of cAMPS-Rp were diluted 1:1 with an intracellular-analogous solution (IAS) which contained 20 mM NaCl, 130 mM KCl, 2 mM MgCl 2 and 5 mM HEPES at pH 7.3. This mixture (9.2 nl) was injected into oocytes, leading to concentrations of ~87 µM cAMPS-Rp per oocyte. Corresponding control experiments were performed with IAS. The membrane permeable cAMP-analogue 8-Br-cAMP (Tocris Bioscience) was prepared as a 10 mM stock solution in H 2 O and stored at −20 °C.
SCIeNTIfIC RepoRts | 7: 3517 | DOI:10.1038/s41598-017-03742-5 Drug application. Drugs were generally applied using a gravity-driven perfusion system. In order to reduce the amount of drugs needed, the membrane-permeable cAMP-analogue 8-Br-cAMP was washed into the perfusion chamber and the perfusion was stopped afterwards. The PKA inhibitor cAMPS-Rp was directly injected into the oocytes since this strategy was established earlier and achieved adequate inhibition of PKA 55 .
Statistical analysis. For electrophysiological transmembrane recordings, outward-anion currents are defined as negative current signals and depicted in all figures as downward-deflections of the current traces. Data are presented as individual data points (grey symbols) as well as means ± standard error of the mean (SEM). The number of individual oocytes is indicated with 'n' , whereas the number of donor frogs is represented by 'N' . Statistical analysis of data was performed with GraphPad Prism version 5 (La Jolla, California, USA). Data were analysed for normal distribution by the Kolmogorov-Smirnov test. For paired experiments, Student's paired t-test or non-parametric Wilcoxon matched-pairs test (two-tailed) were used. Independent experiments were compared with Student's unpaired t-test or non-parametric Mann-Whitney test (two-tailed). Mutiple comparison analysis was performed by Kruskall-Wallis test followed by Dunn's Multiple Comparison Test. P-values ≤ 0.05 were regarded as statistically significant and marked with an asterisk (*). P-values ≤ 0.01 and ≤0.001 were marked with "**" and "***", respectively.