The firing frequency of spontaneous action potentials and their corresponding evoked exocytosis are increased in chromaffin cells of CCl4‐induced cirrhotic rats with respect to control rats

High catecolamine plasma levels because of sympathetic nervous system over‐activity contribute to cirrhosis progression. The aim of this study was to investigate whether chromaffin cells of the adrenal gland might potentiate the deleterious effect exerted by this over‐activity. Electrophysiological patch‐clamp and amperometric experiments with carbon‐fibre electrodes were conducted in single chromaffin cells of control and CCl4‐induced cirrhotic rats. The spontaneous action potential firing frequency was increased in chromaffin cells of cirrhotic rats with respect to control rats. The exocytosis evoked by that firing was also increased. However, exocytosis elicited by ACh did not vary between control and cirrhotic rats. Exocytosis triggered by depolarizing pulses was also unchanged. Amperometric recordings confirmed the lack of increased catecholamine charge released in cirrhosis after ACh or depolarization stimuli. However, the amperometric spikes exhibited faster kinetics of release. The overall Ca2+ entry through voltage‐dependent Ca2+ channels (VDCC), or in particular through Cav1 channels, did not vary between chromaffin cells of control and cirrhotic rats. The inhibition of VDCC by methionine‐enkephaline or ATP was not either altered, but it was increased by adrenaline in cells of cirrhotic rats. When a cocktail composed by the three neurotransmitters was tested in order to approach a situation closer to the physiological condition, the inhibition of VDCC was similar between both types of cells. In summary, chromaffin cells of the adrenal gland might contribute to exacerbate the sympathetic nervous system over‐activity in cirrhosis because of an increased exocytosis elicited by an enhanced spontaneous electrical activity.

Cirrhosis is a pathologically defined entity associated to a spectrum of characteristics clinical manifestations, which should be viewed as a common pathway of many types of chronic liver injury. Regardless of the cause of cirrhosis, the pathologic features consist of the development of fibrosis and architectural distortion with the formation of regenerative nodules. Eventually, cirrhosis progresses to hepatic insufficiency, complications of portal hypertension and death.
Several lines of evidence indicate that sympathetic nervous system (SNS) over-activity develops in the advanced stages of cirrhosis and contributes to disease progression. Activation of the SNS with increased plasma levels of noradrenaline is well documented in patients and experimental models with cirrhosis and ascites (Henriksen et al. 1984(Henriksen et al. , 1985(Henriksen et al. , 1987(Henriksen et al. , 1988(Henriksen et al. , 1998Nicholls et al. 1985;Bernardi et al. 1987;Floras et al. 1991;Esler et al. 1992;Iwao et al. 1994;Kirstetter et al. 1996;Pozzi et al. 2001;Sanso e et al. 2016). SNS over-activity participates in cirrhosis progression by exacerbating liver fibrosis and by worsening portal hypertension. Catecholamines contribute to fibrosis progression by stimulating proliferation and collagen gene expression (Oben et al. 2003a(Oben et al. ,b,c, 2004a, and by exerting pro-inflammatory actions on hepatic stellate cells (HSCs) (Sancho-Bru et al. 2006). In this regard, liver fibrosis is more severe in spontaneously hypertensive than in Wystar-Kyoto rats (Hsu 1992), and CCl 4 -induced liver fibrosis is attenuated after sympathectomy with 6-hydroxydopamine or a 1 -adrenoceptor antagonism with prazosin (Dubuissons et al. 2002). In addition, catecholamines worsen portal hypertension in cirrhosis by increasing intrahepatic vascular resistance to blood flow (Ballet et al. 1988;Esler et al. 1992). In support of this concept, a-adrenergic block with prazosin or sympathetic nervous system outflow inhibition with clonidine lower portal pressure in cirrhosis (Albillos et al. 1992(Albillos et al. , 1994. The activation of the SNS presents a regional distribution in cirrhosis with a predominant activity in the splanchnic area (Henriksen et al. 1988;Pozzi et al. 2001). Furthermore, down-regulation of genes related to the adrenergic system such as decreased protein expression of tyrosine-hydroxylase, the key enzyme for noradrenaline and adrenaline synthesis (Coll et al. 2008) and atrophy of mesenteric sympathetic innervation (Coll et al. 2010), have been reported and may contribute to splanchnic vasodilation in rats with portal hypertension.
It has been posed by Oben and Diehl (2004) that 'The in vivo sources that might provide noradrenaline for HSC regulation include HSCs themselves, SNS nerve terminals that about HSCs, and the adrenal medulla that releases noradrenaline and adrenaline into the circulation under stressful conditions, such as liver injury. The relative importance of these three sources in the regulation of HSC function in vivo is as yet unclear'. Thus, the relevance of SNS over-activity in cirrhosis has fostered the interest to identify whether other sources of catecholamines, such as the chromaffin cells of the adrenal gland, may potentiate the SNS deleterious effect in this disease. We aimed at investigating in chromaffin cells of the adrenal gland of rats with CCl 4 -induced cirrhosis whether exocytosis elicited at basal conditions because of the spontaneous firing of action potentials, or exocytosis triggered by different stimuli (ACh or depolarizing pulses), might be altered in cirrhosis. We also investigated whether Ca 2+ currents, flowing through voltagedependent Ca 2+ channels (VDCC) or their modulation by neurotransmitters, were modified in chromaffin cells of the adrenal gland of cirrhotic rats in comparison to control rats.

Animals
Male Wistar Hannover rats (RccHan:WIST) (Envigo, Horst, The Netherlands, cat. #168) of 20 to 31 weeks old were used for all the experiments ( Fig. 1). They had an average weight of 385 g. Animals were fed with a standard laboratory diet with water and food provided ad libitum. The study was not pre-registered. Experiments were approved by and performed in accordance with the Ethic Committee of the Universidad Aut onoma de Madrid and Universidad de Alcal a regulations and conducted according to the European Directive 2010/63/EU and Royal Decree 53/2013 from Spain. No randomization method was employed to allocate animals to different experimental groups.

Induction of cirrhosis
Cirrhosis was induced by CCl 4 (Sigma Aldrich, Madrid, Spain, cat. #289116), feeding by gavage on a weekly basis, along with phenobarbital (Qu ımica Farmac eutica Bayer, Barcelona, Spain) bought in a local pharmacy, added to the drinking water (0.35 g/L). The initial 20 lL dose of CCl 4 was subsequently increased, depending on the animal weekly change in body weight until ascites formation (Runyon et al. 1991). Experiments were performed 7 days after the last CCl 4 dose ( Fig. 1). Cirrhosis was confirmed by trichrome staining of livers. Only rats with ascites were included in the study (n = 18). Two rats with cirrhosis but without ascitis were excluded from the study. Phenobarbital-treated age-and sex-matched rats were used as the control group (n = 16). No statistical methods were performed to predetermine the sample size. Barbiturates have been reported to inhibit nicotinic acetylcholine receptors in a reversible manner (Watanabe et al. 1999), and therefore, phenobarbital should not affect the response to ACh in these experiments.
Isolation and culture of rat chromaffin cells Animals were killed inside a gas chamber and were opened along the peritoneal cavity. To obtain the glands, tweezers of sharp tip and scissors were employed. Once removed, they were placed on a 35 mm diameter Petri dish with Locke solution (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO 3 , 5 mM HEPES and 5.6 mM glucose,) maintained on ice. Under the microscope, fat tissue and cortex were taken off by means of tweezers and scalpel on the Petri dish with ice. The medullae were introduced in Locke solution, and afterwards they were incubated during 60 min in control rats, or 50 min in cirrhotic rats, in a solution with trypsin (2.5%) (GIBCO, Invitrogen Life Technologies, Massachusetts, USA, cat. #15090-046) and collagenase (1%) (Sigma-Aldrich, cat. #C0130) at 37°C. After this time they were washed with Locke, and later on with Dulbecco's Modified Eagle's Medium (Sigma-Aldrich, cat. #D6546) that was supplemented with 1% penicillin-streptomycin (Sigma-Aldrich, cat. #P4333), 1% GlutaMAX (GIBCO, cat. #35050038) and 5% foetal bovine serum (Labclinics, cat. #S181B). Mechanic digestion was performed by passing the medullae through a pipette tip until obtaining an homogeneous suspension. The cells were plated on glass coverslips previously treated with polylysine (0.1 mg/mL) (Sigma-Aldrich, cat. #P2636). Finally they were introduced in an incubator with saturated atmosphere of water vapour, 95% of O 2 and 5% of CO 2 . Experiments were performed 1-4 days after plating on matched time periods.
An amphotericin B (Sigma-Aldrich, cat. #A4888) stock solution was prepared every day at a concentration of 50 mg/mL in dimethyl sulphoxide and kept protected from light. The final concentration of amphotericin B was prepared by ultrasonicating 10 lL of stock amphotericin B in 1 mL of Cs-glutamate internal solution in the dark. Pipettes were tip-dipped in amphotericin-free solution for several seconds and back-filled with freshly mixed intracellular amphotericin solution.
Chemical compounds were purchased from Sigma-Aldrich. Experiments were performed at room temperature (22-24°C). There were no sample size differences between the beginning and end of the experiments. No blinding procedures were performed.
Drugs were applied by means of a multi barrelled pipette that was constructed using polyethylene tubing with an inner diameter of 0.4 mm. These tubes coalesced to a single outlet tube with a 0.28 mm inner diameter. Drugs were delivered by gravity and were controlled by a valve controller triggered by the amplifier. A Picospritzer III (General Valve Corp., Fairfield, NJ, USA) was used to apply ACh to the cell by means of pressure ejection (15 psi) through a glass capillary tube (World Precision Instruments, Sarasota, FL, USA; cat. 1B200F-4) that was pulled to obtain an opening of 1-2 lm in diameter. For a rapid wash-out of ACh, a pipette with a polyethylene tube with an inner diameter of 0.58 mm was used. The outlet of this tube was placed just behind, close to the glass capillary tube of the Picospritzer.
Electrophysiological measurements were made using an EPC-10 amplifier and PULSE software (HEKA Elektronik, Lambrecht, Germany) running on a PC computer. Glass electrodes were pulled from borosilicate glass capillaries (Kimbal Chase, cat. #3400-99) using a P97 pipette puller (Sutter Instruments, Novato CA, USA). These electrodes had resistances between 1.5 and 3 MΩ when they were filled with the internal electrode solutions. Borosilicate glass capillary tubes were partially coated with wax and fire polished. Only recordings in which the leak current and access resistance were lower than 20 pA and 25 MΩ in the perforated-patch configuration, Week 5 Week 20-31 Sacrifice → cell culture Week 19-30 Week 19-30 respectively, were accepted. We noticed more difficulty in perforating the chromaffin cell membrane of the 5-7 months old rats of this study with respect to younger rats. Cell membrane capacitance (C m ) changes as an index of exocytosis were estimated by the Lindau-Neher technique implemented in the 'Sine+DC' feature of the 'PULSE' lock-in software. A 1 kHz, 70 mV peak-to-peak amplitude sinewave was applied at a holding potential (V h ) of À80 mV. The signals were sampled at 10 kHz and filtered at 1 kHz through a Bessel filter. Cells were clamped at a V h of À80 mV with the exception of those experiments performed using ACh as a stimuli and the 'triple-step' protocol to measure exocytosis (P erez-Alvarez and Albillos 2007), which were carried out at the resting membrane potential of each individual cell. Chromaffin cells of control or cirrhotic rats exhibited the same specific capacitance, 7.8 AE 0.4 pF (n = 70) and 7.1 AE 0.3 pF (n = 77), respectively, reflecting a similar size. Analysis of electrophysiological data was conducted using IGOR Prosoftware (RRID: SCR_000325, Wavemetrics, Lake Oswego, Oregon). The ROUT test of GraphPad Prism was used to identify outliers. In the analysis of the inhibition by ATP of the Ca 2+ current charge, 1 of 9 control cells and 1 cell of 10 cirrhotic cells, respectively, were excluded from the analysis (they potentiated the current by 2.2% and 22.1% respectively). In the case of the inhibition of VDCC by adrenaline, 1 of 13 control cells was excluded from the analysis (it potentiated the current by 84%). In addition, 1 of 17 control cells and 1 of 12 cirrhotic cells showed potentiation in the perforated-patch clamp configuration after perfusion with a cocktail of methionine-enkephalin, ATP and adrenaline (4.3% and 9.5%, from control and cirrhotic rats respectively). These cells were not either included in the analysis of the inhibition of VDCC by the cocktail of compounds.
In the triple step protocol, the C m increment was calculated from the baseline in Step 1 to the initial value of C m recorded in Step 3, once conductances activated in Step 2 returned to the zero value. In the case of C m increments because of depolarizing pulses, they were calculated from the baseline C m recorded before the depolarizing pulse to the maximum value obtained after the pulse, once activated conductances during the depolarization returned to zero. Inhibition of Ca 2+ currents were calculated as the decrease between the peak control current value and the corresponding current value in the presence of the drug under study. The nonspecific background current and C m recorded under 200 lM CdCl 2 were subtracted offline from Ca 2+ current and C m traces. Protocols were repeated every 5 min in order to obtain an average value of two or three measurements of exocytosis.
Amperometric recordings and analysis of data Carbon fibre electrodes were prepared by cannulating a 10 lmdiameter carbon fibre in polyethylene tubing (diameter: outer, 1 mm; inner, 0.5 mm). The carbon fibre tip was glued into a glass capillary for mounting on a patch-clamp headstage, and back filled with 3 M KCl to connect to the Ag/AgCl wire, which was kept at +700 mV. The carbon fibre electrode was gently placed on top of the cell under study. Amperometric currents were recorded using an EPC-10 amplifier and PULSE software running on a PC computer. The sampling rate was 14.5 kHz. Samples were digitally filtered at 0.1 kHz. The sensitivity of the electrodes was routinely monitored before and after the experiments using 50 lM of adrenaline as standard solution. Only fibres that rendered 200-300 pA of current increment after 50 lM of adrenaline pulse were used for the experiments. The tip of the fibre was recut for each experiment and calibrated again.
Chromaffin cells were perfused with a solution containing (in mM): 145 NaCl, 1 MgCl 2 , 10 HEPES, 5.5 KCl, 2 CaCl 2 and 10 glucose (pH 7.4). Cells were stimulated with 300 lM ACh or 100 mM K + . There were no sample size differences between the beginning and end of the experiments. No blinding procedures were performed.
Spike analysis was performed using IGOR Pro software and macros that allowed the analysis of single events and the rejection of overlapping spikes (Mosharov and Sulzer 2005). A threshold of 4.5 times the first derivative of the noise standard deviation was calculated to clearly detect amperometric events. Then, among the events whose first derivative was above this threshold, only those showing one peak and one rising and falling phase, were considered as single spikes.
Statistical analysis of data Statistical analysis was performed using SPSS 24.0 (RRID: SCR_002865, Armonk, NY, USA). Data were given as the mean AE SEM for the number (n) of cells. The normality test Kolmogorov-Smirnov was first performed. Data were then compared using the unpaired Student's t-test or Mann-Whitney U test. Data were found statistically significant when *p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001.

Results
Exocytosis elicited by the firing of spontaneous action potentials The exocytosis or fusion process of chromaffin vesicles with the plasma membrane was determined by recording the increments in plasma membrane capacitance (C m ) (Lindau and Neher 1988;Albillos et al. 1997). We first investigated whether exocytosis because of the firing of spontaneous action potentials (sAP) at rest could be responsible for the increased basal catecholamine plasma levels previously reported in cirrhosis. Chromaffin cells fire sAP (Biales et al. 1976;Marcantoni et al. 2007Marcantoni et al. , 2009Marcantoni et al. , 2010P erez-Alvarez et al. 2011;Hern andez-Vivanco et al. 2017), but it has not been probed so far that these sAP can evoke the fusion of secretory vesicles with the plasma membrane, although catecholamine release has been determined (Vandael et al. 2015a). Here we measured for the first time the exocytosis because of the firing of sAP using the 'triple-step' protocol previously performed in our laboratory (P erez-Alvarez and Albillos 2007). It consists in the successive switch from the voltage-clamp configuration to determine basal C m (Step 1), to the current-clamp configuration to allow sAP firing (Step 2), and then again to the voltage-clamp configuration to measure the increased C m because of the previous firing of sAP (Step 3). A scheme of the protocol is displayed in Fig. 2a. This protocol was repeated every 5 min in order to Step 2), and then switch again to the VC configuration (Step 3) to record the increment in C m elicited by the sAP recorded in Step 2. V h was the resting membrane potential of each particular cell. This protocol was performed in control n = 5 cells (panel b) and cirrhotic n = 7 cells (panel c) from three independent cultures. It was repeated two-three times every 5 min. The time course of firing is showed in the separate panels corresponding to the grey windows above.
obtain an average value of two or three measurements of exocytosis. The V h was the resting membrane potential of each cell, which was similar between chromaffin cells of control and cirrhotic rats. The resting membrane potential was obtained 5 min after establishing the current-clamp configuration. Values were À52.5 AE 2.6 mV in control rats (n = 8) and -54.6 AE 3.5 mV in cirrhotic rats (n = 13). The sAP firing frequency was 1.4 AE 0.4 Hz (n = 5) and 2.6 AE 0.3 Hz (n = 7) (*p ≤ 0.05), and the exocytosis achieved was 46.8 AE 13 fF (n = 5) and 160.3 AE 31.1 fF (n = 7) in chromaffin cells of control and cirrhotic rats respectively (**p ≤ 0.01). Representative recordings of this type of experiment are shown in Fig. 2b and c for control and cirrhotic rats respectively. The time course of firing is showed into more detail in the grey separate panels on an expanded time scale. The half-width (duration of the sAP at 50% of its peak amplitude) and amplitude (maximal potential achieved from the basal potential) of the sAP were unchanged between control and cirrhotic rats. Values amounted to 6 AE 0.7 ms and 5.6 AE 0.3 ms for the halfwidth and 50.6 AE 2.6 mV and 46.4 AE 3.7 mV for the amplitude of the sAP in control (n = 5) and cirrhotic rats (n = 7) respectively.

Exocytosis evoked by short pulses of ACh or depolarizing pulses
Plausible changes in relation to the stimulated exocytosis were also evaluated by applying two types of stimuli in control rats in comparison to cirrhotic rats: ACh, the physiological neurotransmitter at the synapse chromaffin cell-splanchnic nerve, and depolarizing stimuli. In the first case, 10 ms pulses of 300 lM concentration ACh (11 pulses) at 0.2 Hz were applied at the resting membrane potential of each cell. That concentration of ACh was chosen for two reasons. First, in a post-synaptic area of 1 mm 2 with a 50-nm-wide synaptic cleft, the peak ACh concentration is 0.3 mM (Kuffler and Yoshikami 1975;Scimemi and Beato 2009). Second, 300 lM is the concentration of ACh that elicits larger nicotinic currents with the minimum openchannel block effect in human chromaffin cells (P erez-Alvarez and Albillos 2007; P erez-Alvarez et al. 2012a,b; Hone et al. 2015Hone et al. , 2017, and also in rat chromaffin cells (Hone et al., unpublished data). We tested a very short duration of ACh pulses (10 ms) applied at the resting membrane potential of the cell and at the basal firing frequency of the splanchnic nerve, to mimick conditions closer to the physiological situation. The overall exocytosis evoked by ACh was also determined by using the 'triplestep' protocol (increment between the C m of Step 3 and 1) (Fig. 3a). Changes in the membrane potential because of the application of 10 ms ACh pulses were recorded in the current-clamp configuration (Step 2). ACh evoked 'Spikelike' changes in the membrane potential. The triple-step protocol was repeated every 5 min to obtain an average value of exocytosis. The exocytosis triggered by the ACh pulses was 278.1 AE 76 fF (n = 8) and 272.6 AE 86 fF (n = 6) in control and cirrhotic rats respectively. Representative recordings of this protocol are shown in Fig. 3b and c for control and cirrhotic rats respectively.
In a different set of experiments, depolarizing pulses of 200 ms to the peak current voltage were applied every 5 min. In order to know the voltage at which Ca 2+ current was maximal, a ramp test of 200 ms from -100 mV to 100 mV was applied at the beginning of the experiment. The increment achieved in C m was 71 AE 12.4 fF (n = 13) and 95.8 AE 20 fF (n = 11), in control and cirrhotic rats respectively. No significant statistical differences were found between both groups of rats. Endocytosis was also similar between cells of control and cirrhotic rats, amounting to 71.7 AE 21.3 fF (n = 13) and 79.1 AE 23.8 fF (n = 11) respectively. Some cells did not exhibit endocytosis, whereas other cells showed a variable degree of compensatory endocytosis in both groups of cells. Representative C m recordings are displayed in Fig. 3d and e for control and cirrhotic rats respectively.

Release of catecholamines in cells stimulated with ACh or K +
In order to confirm the data obtained by C m measurements in relation to the trigger of the exocytosis by ACh or depolarization, the release of catecholamines elicited by activation of nAChRs using ACh as stimulus, or by depolarization of the plasma membrane using 100 mM K + , was now recorded using a carbon fibre electrode. The transient current of oxidation because of the catecholamines oxidised at the surface of the electrode, at +700 mV potential, was recorded and compared in chromaffin cells of control and cirrhotic rats. Representative recordings of catecholamine release of single chromaffin cells stimulated with 2 s pulses of 300 lM ACh are displayed from control and cirrhotic rats ( Fig. 4a and b respectively) or 100 mM K + ( Fig. 4c and d respectively). The charge corresponding to the overall catecholamines released was similar for both stimuli between control and cirrhotic rat chromaffin cells. The values achieved were 25.3 AE 5 pC (n = 13) and 39 AE 10 pC (n = 13) for control and cirrhotic ACh-treated cells, respectively, and 37.4 AE 6.3 pC (n = 17) and 46.3 AE 8 pC (n = 14) for control and cirrhotic K + -treated cells, showing no significant statistical differences.
A detailed analysis of the amperometric spikes obtained was performed. The following parameters were determined: I max , peak amplitude; Q, charge; m, ascending slope; rise 25-75, time calculated from the linear portion of the ascending trace between 25% and 75% of the I max ; fall 75-25, time calculated from the linear portion of the descending trace between 75% and 25% of the I max ; base, time between the spike beginning and end respectively; and t 1/2 , half-width or duration of the amperometric signal at 50% of its peak amplitude. A scheme of these parameters is shown in Fig. 4e. Values were shown in bar diagrams for control (black) and cirrhotic (grey) rats in Fig. 4f and g, for ACh and K + respectively. It was found that the amperometric spikes evoked by ACh or K + in cirrhotic rats exhibited larger I max and m values, and smaller rise 25-75, fall 75-25, base or t 1/2 values (Table 1). This reflects that catecholamines were released faster in chromaffin cells of cirrhotic rats, although the total charge remained unchanged.
Ca 2+ entry charge flowing through VDCC and their modulation by opioids, ATP and adrenaline Depolarizing pulses of 50 ms were applied at the voltage at which Ca 2+ current was maximal in the perforated-patch configuration. To know the value of this voltage, a ramp test of 200 ms from À100 mV to 100 mV was applied at the beginning of each experiment. The total Ca 2+ charge recruited by these depolarizing pulses was 11.1 AE 1 pC (n = 16) and 10.7 AE 1.5 pC (n = 11) in control and cirrhotic rats respectively. No significant statistical differences were found. Therefore, it was not expected that Cav1 channel expression would have been modified, as reported in activated HSCs in comparison with quiescent HSCs (Bataller et al. 2001). Anyway, the contribution of Cav1 channels to the overall Ca 2+ current was evaluated using 3 lM nifedipine to isolate Cav1 channels by subtracting the current in the presence of the drug from the control current, as performed in previous studies The duration of the pulses was 10 ms, applied using a Picotspritzer. Representative recordings of the three steps of the protocol obtained in control n = 8 cells and cirrhotic n = 6 cells from four independent cultures are shown in panels b and c respectively. V h was the resting membrane potential of each cell. The protocol was repeated two-three times every 5 min. (d and e) Representative C m recordings elicited by depolarizing pulses of 200 ms applied from a V h of À80 mV to the voltage that exhibited a larger peak of Ca 2+ current (usually 0 or +10 mV) in control n = 13 cells (panel d) and cirrhotic n = 11 cells (panel e) from four independent cultures. (Albillos et al. 1994(Albillos et al. , 1996aUlate et al. 2000;Aldea et al. 2002;P erez-Alvarez et al. 2008P erez-Alvarez et al. , 2011Marcantoni et al. 2009Marcantoni et al. , 2010Hern andez-Vivanco et al. 2012. Ca 2+ currents were elicited by 100 ms depolarizing pulses. Nifedipine block of the Ca 2+ charge was similar in control and cirrhotic rats, amounting to 44.4 AE 6% (n = 9) and 42.8 AE 3.6% (n = 14) respectively. Representative recordings of Ba 2+ current traces before and after perfusion with nifedipine are shown in Fig. 5a and b, respectively, for control and cirrhotic rats.
However, modulation of VDCC by the neurotransmitters released by chromaffin cells might be impaired in cirrhotic rats, modifying the amount of Ca 2+ entry to the cytosol. This would lead to impaired exocytosis and catecholamine release. The block of methionine-enkephalin, ATP and adrenaline was analyzed separately using the conventional whole-cell configuration of the patch-clamp technique and 10 mM Ba 2+ as cation charger in control and cirrhotic rats. This configuration of patch-clamp was employed in order to introduce GTP inside of the cell, avoiding a decrease in the modulatory effect by the neurotransmitters because of GTP consumption along the experiment. The protocol consisted of depolarizing pulses of 50 ms from the V h to 0 or +10 mV, applied every 15 s. The three neurotransmitters were used at 10 lM concentration (Albillos et al. 1996a and b;Gand ıa et al. 1993). Methionine-enkephaline inhibited by 21.5 AE 5% (n = 12) and 15.8 AE 4.1% (n = 9) the Ba 2+ charge in control and cirrhotic rats respectively ( Fig. 5c and  d). ATP inhibited by 24.6 AE 9% (n = 8) and 27.4 AE 10.4% (n = 9) the Ba 2+ charge in control and cirrhotic rats respectively. Representative current recordings before and after perfusion ATP are displayed in Fig. 5e and f, for control and cirrhotic rats respectively. Finally, adrenaline inhibited by 14 AE 3.6% (n = 12) and 35 AE 6% (n = 6) the Ba 2+ charge in control and cirrhotic rats respectively (**p ≤ 0.01). Representative current recordings before and after perfusion adrenaline are shown in Fig. 5g and h, respectively, for control and cirrhotic rats.
In order to know whether differences might exist when neurotransmitters were applied together, a condition closer to the physiological situation, the effect of a cocktail of Fig. 4 Catecholamine release evoked by 5 s pulses of ACh (300 lM) or K + (70 mM) using carbon-fibre electrodes. (a and b) Amperometric recordings performed in control n = 13 cells and cirrhotic n = 13 cells from six independent cultures, stimulated with ACh. (c and d) Amperometric recordings performed in control n = 17 cells and cirrhotic n = 14 cells from six independent cultures, stimulated with K + . (e) Scheme of the analyzed kinetic parameters of the amperometric spikes. These parameters include the peak amplitude (I max ), measured between the current at its maximum value and the baseline current under the spike maximum; the base (Base), is the time between the spike beginning and end respectively; spike width or duration of the amperometric signal at 50% of its peak amplitude (t 1/2 ), is evaluated at 50% of I max ; charge, or integral of individual amperometric spikes (Q), estimated from spike area above the baseline; rising phase or slope (m) is calculated using a linear fit to the current between two points on the ascending segment of the spike; rise 25-75 is the time from the 25% to the 75% of I max ; and fall 75-25 is the time from the 75% to the 25% of I max . (f) Diagram bars of the parameters obtained from the amperometric spikes elicited by ACh in control (black) and cirrhotic (grey) chromaffin cells. (g) Diagram bars and single data points showed as an overlaying aligned dot plot of the parameters obtained from the amperometric spikes elicited by K + in control (black) and cirrhotic (grey) chromaffin cells. *p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001. methionine-enkephalin, ATP and adrenaline was investigated on VDCC in the conventional whole-cell configuration using 10 mM Ba 2+ as charge carrier in control and cirrhotic rats. The block of Ca 2+ charge achieved by the cocktail of neurotransmitters was 37.2 AE 5% (n = 18) and 50.5 AE 7% (n = 12) in control and cirrhotic rats respectively. No significant statistical differences were obtained. Representative recordings before and after perfusion the cocktail of compounds are shown in Fig. 5i and j respectively. We wanted to confirm the results obtained in the conventional whole-cell configuration using the whole-cell perforatedpatch configuration (in 5 mM Ca 2+ as charge carrier) which avoids dilution of intracellular factors. Using this condition, the cocktail of compounds inhibited by 42.1 AE 8.1% (n = 16) and 39.4 AE 6% (n = 11) the Ca 2+ charge in control and cirrhotic rats. No significant statistical differences were obtained. Representative recordings before and after perfusion the cocktail of compounds are displayed in Fig. 5k and l, respectively, for control and cirrhotic rats.

Discussion
Cirrhosis is characterized by a SNS over-activity with elevated noradrenaline and adrenaline plasma levels that contribute to the progression of the disease. It has been postulated that other sources of catecholamines different to sympathetic nerve fibres, such as chromaffin cells of the adrenal gland medulla, may also regulate HSCs (Oben and Diehl 2004). Elevated catecholamines plasma levels have been reported in cirrhotic patients (Henriksen et al. 1984(Henriksen et al. , 1985(Henriksen et al. , 1987(Henriksen et al. , 1988(Henriksen et al. , 1998Nicholls et al. 1985;Iwao et al. 1994;Pozzi et al. 2001). In addition, also patients suffering essential hypertension (Goldstein et al. 1983) or spontaneously hypertensive rats (SHR) (Iriuchijima 1973;Grobecker et al. 1975) exhibit high catecholamine plasma levels. In this latter case, greater catecholamine release was observed, as compared with normotensive rats (Lim et al. 2002). Indeed, SHR chromaffin cells showed faster and larger catecholamine responses, explained by more vesicles ready to undergo exocytosis and greater quantal content of vesicles (Miranda-Ferreira et al. 2008) rather than altered Ca 2+ entry and subsequent redistribution into the endoplasmic reticulum or mitochondria (Miranda-Ferreira et al. 2009). Based on these results, our hypothesis posed that impaired Ca 2+ entry to the cytosol through VDCC or altered spontaneous actions potentials (sAP) in chromaffin cells might lead to a larger exocytosis and catecholamine release, that would contribute to exacerbate the negative effects of the SNS over-activity reported in liver cirrhosis. Similarly to what it was obtained in SHR chromaffin cells, amperometric spikes recorded using ACh or high concentration of K + showed faster catecholamine responses in cirrhotic with respect to control chromaffin cells. However, the quantal content of the vesicles remained unchanged. Chromaffin cells of the adrenal gland are modified postganglionic sympathetic neurons innervated by the splanchnic nerve that mainly control the release of adrenaline to the bloodstream, to prepare muscle and cardiovascular systems to a situation of stress. In humans, most of chromaffin cells possess an adrenergic phenotype (P erez-Alvarez et al. 2008). Chromaffin cells store a pleyade of compounds which include catecholamines at concentrations as high as 1 M, opioids, ATP, chromogranins (components of the vesicular matrix) or Ca 2+ in chromaffin vesicles, which are dense core granules that fuse with the plasma membrane through a Ca 2+dependent mechanism called exocytosis. This process is mediated through SNAREs (Soluble N-ethylmaleimidesensitive factor Attachment Protein Receptors) and Rab family proteins. During exocytosis, a narrow fusion pore is formed, allowing the release of free catecholamines, and afterwards, this pore expands, releasing the whole content of the vesicle ('full fusion event'). Alternatively, the opening of the fusion pore can be transient, allowing only the release of catecholamines but not of the vesicular matrix proteins ('kiss and run event') (Albillos et al. 1997). The transport of secretory vesicles, the docking of these vesicles with the plasma membrane, as well as the regulation of the expansion of the fusion pore, are also Ca 2+ -dependent mechanisms. Therefore, regulation of exocytosis and the release process can be performed through regulation of Ca 2+ entry, which would include, among other targets, VDCC.
We evaluated the exocytosis yielded by the firing of sAP in these cells by measuring the C m increments owing to that firing. In mouse chromaffin cells, the catecholamine released by the firing of sAP was previously recorded using the carbon fibre amperometry in combination with current-clamp recordings (Vandael et al. 2015a). We achieved that measurement by using the 'triple-step' protocol previously Fig. 5 Cav1 channels and modulation of voltage-dependent calcium channel (VDCC) in chromaffin cells of control and cirrhotic rats. (a) and (b) Ca 2+ currents elicited by 100 ms depolarizing pulses, from a V h of À80 mV to the peak current voltage, recorded in the perforated-patch configuration of the patch-clamp technique under control conditions (black trace) or after perfusion 3 lM nifedipine (grey trace) in control n = 9 cells and cirrhotic n = 14 cells from six independent cultures. (c and d) Ba 2+ currents elicited by 50 ms depolarizing pulses, from a V h of À80 mV to the peak current voltage, recorded in the whole-cell configuration of the patch-clamp technique under control conditions (black trace) or after perfusion of 10 lM methionine-enkephaline (grey trace), in control n = 12 cells and cirrhotic n = 9 cells from seven independent cultures. The same type of experiments were performed to test 10 lM ATP (e and f) (control n = 8 cells and cirrhotic n = 9 cells from nine independent cultures), 10 lM adrenaline (g and h) (control n = 12 cells and cirrhotic n = 6 cells from eight independent cultures), a cocktail of compounds composed by 10 lM methionine-enkephaline, 10 lM ATP and 10 lM adrenaline (i and j) (control n = 18 cells and cirrhotic n = 12 cells from ten independent cultures), and the same cocktail of compounds tested in the perforated-patch configuration (k and l) (control n = 16 cells and cirrhotic n = 11 cells from six independent cultures). described in our laboratory that allows to determine C m increments because of the changes in membrane potential (P erez-Alvarez and Albillos 2007). We obtained that the frequency of sAP firing in chromaffin cells of cirrhotic rats was increased with respect to control rats, and this elicited a higher exocytotic response. It has been shown that the firing mode can influence secretion in chromaffin cells (Duan et al. 2003;Vandael et al. 2015a;Guarina et al. 2017). However, in this study, the firing frequency of action potentials and the corresponding exocytosis increased without changing the firing mode.
Many ion channel types regulate chromaffin cell excitability, and therefore, might be responsible for the increased firing frequency. In chromaffin cells, Cav1 channels fire spontaneous action potentials (Marcantoni et al. 2007(Marcantoni et al. , 2010P erez-Alvarez et al. 2011). They express two Ca 2+ channel subtypes, Cav1.2 and Cav1.3, which open at relatively low membrane potentials and allow Ca 2+ to enter the cells near resting potentials (Marcantoni et al. 2010;P erez-Alvarez et al. 2011). In this way, these channels shape the action potential waveform and pacemaker activity. Besides that, Cav1.3 supports most of the pacemaking current that sustains action potential firings and part of the catecholamine secretion (Vandael et al. 2015b). In cirrhotic rats a higher expression of Cav1 channels in HSCs has been shown (Bataller et al. 2001), which would allow larger Ca 2+ entry to the cytosol and therefore, Ca 2+ dependent-exocytosis and neurotransmitter or collagen release. Therefore, a larger expression of Cav1 channels, in particular of Cav1.3, might explain the increased frequency of firing in chromaffin cells of cirrhotic rats reported in this study. However, nifedipine block was similar in both types of rats, which indicates that a similar amount of Cav1 channels are expressed. Anyway, we can not exclude that Cav1.3 channels are more expressed in cirrhotic rats, as the dihydropyridine blocks both channel subtypes.
A strong Cav1.3-BK channels coupling in wild-type mouse chromaffin cells as well as in rat chromaffin cells have been also reported, although rat chromaffin cells express higher densities of BK channels (Prakriya and Lingle 1999;Marcantoni et al. 2010). These channels are activated by both the action potential and Ca 2+ entering the cytoplasm during the interspike, which is mainly carried by Cav1.3 channels. BK currents sustain mainly the afterhyperpolarization of the short action potential and only partially the pacemaker current during the long interspike in mouse chromaffin cells (Marcantoni et al. 2010). In addition, a Cav1.3-driven SK channel activation regulates pacemaking and spike frequency adaptation in mouse chromaffin cells (Vandael et al. 2012). Other channels such as Na + (Nav1.3 and Nav1.7) or K + channels (Kv1-3, Kv4, Kv7, Kv11 and K2P), also contribute to regulate chromaffin cell excitability and impact on it under pathological circumstances (Lingle et al. 2018). Further research would be required to clarify the ion channel conductances altered in chromaffin cells in cirrhosis.
In relation to the regulation of VDCC, these can be modulated by the products released through a negative feedback mechanism in bovine chromaffin cells (Albillos et al. 1996a and b). In particular, ATP (Gand ıa et al. 1993), opioids (Albillos et al. 1996a) and adrenaline (Albillos et al. 1996b) inhibit the overall Ca 2+ current recruited by step-wise depolarizing pulses through a modulatory pathway mediated by G proteins. Here we obtained that this modulatory pathway was not impaired in chromaffin cells of the adrenal gland of rats with Cl 4 C-induced cirrhosis in the case of methionine-enkephaline or ATP, but it was increased in the case of adrenaline. This might indicate an adrenergic overactivity in chromaffin cells of cirrhotic rats, which would be in line with the SNS over-activity reported in cirrhosis. However, when methionine-enkephaline, ATP and adrenaline were perfused together, the inhibition was increased to values between 40-50%, and they were similar in both types of cells. This may be because of the fact that all these neurotransmitters act through the same G-protein-mediated pathway, which would be saturated, reaching maximal and identical values in both types of cells. These experiments were performed in the conventional whole-cell configuration of the patch-clamp technique in order to introduce GTP inside of the patch-pipette, to avoid a decrease modulatory effect of the tested neurotransmitters as a consequence of GTP consumption during the experiment. Later on, experiments were also conducted in the perforated-patch configuration, to record the modulation of VDCC by the cocktail of neurotransmitters under more physiological conditions, which might include a differential cytosolic concentration of GTP between control and cirrhotic rats that would lead to a different modulatory effect by neurotransmitters. Using both methods, we obtained that the cocktail of ATP, methionine-enkephaline and adrenaline did not exert a different inhibitory effect on Ca 2+ currents between control and cirrhotic rats.
In summary, our data reveal that in cirrhosis, adrenal gland chromaffin cells of the sympato-adrenal axis might contribute to potentiate the SNS over-activity that worsens this disease by increasing the exocytotic response triggered by an enhanced spontaneous electrical activity.