Argon reduces the pulmonary vascular tone in rats and humans by GABA-receptor activation

Argon exerts neuroprotection. Thus, it might improve patients’ neurological outcome after cerebral disorders or cardiopulmonary resuscitation. However, limited data are available concerning its effect on pulmonary vessel and airways. We used rat isolated perfused lungs (IPL) and precision-cut lung slices (PCLS) of rats and humans to assess this topic. IPL: Airway and perfusion parameters, oedema formation and the pulmonary capillary pressure (Pcap) were measured and the precapillary and postcapillary resistance (Rpost) was calculated. In IPLs and PCLS, the pulmonary vessel tone was enhanced with ET-1 or remained unchanged. IPLs were ventilated and PCLS were gassed with argon-mixture or room-air. IPL: Argon reduced the ET-1-induced increase of Pcap, Rpost and oedema formation (p < 0.05). PCLS (rat): Argon relaxed naïve pulmonary arteries (PAs) (p < 0.05). PCLS (rat/human): Argon attenuated the ET-1-induced contraction in PAs (p < 0.05). Inhibition of GABAB-receptors abolished argon-induced relaxation (p < 0.05) in naïve or ET-1-pre-contracted PAs; whereas inhibition of GABAA-receptors only affected ET-1-pre-contracted PAs (p < 0.01). GABAA/B-receptor agonists attenuated ET-1-induced contraction in PAs and baclofen (GABAB-agonist) even in pulmonary veins (p < 0.001). PLCS (rat): Argon did not affect the airways. Finally, argon decreases the pulmonary vessel tone by activation of GABA-receptors. Hence, argon might be applicable in patients with pulmonary hypertension and right ventricular failure.


Results
We evaluated the effects of argon on pulmonary haemodynamic and airway parameters using rat isolated perfused lungs (IPL) and precision-cut lung slices (PCLS) of rats and humans. In both models, healthy lungs with or without ET-1 pre-treatment were studied. In addition, we addressed the role of GABA-receptor activation within argon-induced relaxation in naïve or ET-1 pre-contracted rat PCLS.
The effects of argon on pulmonary haemodynamics and airway parameters in the IPL. In IPLs, ventilation with argon was started, if baseline parameters were stable for 20 minutes. Control lungs were forward ventilated with room-air. Ventilation with argon did not alter pulmonary haemodynamics (Fig. 1A-E), e.g. pulmonary arterial pressure (P PA ), PVR, P cap , R pre , R post or oedema formation, indicated by the wet-to-dry ratio (W/D-ratio; Fig. 1F). However, ventilation with argon reduced the tidal volume (TV) ( Fig. 2A; p < 0.001) and the lung compliance (C) (Fig. 2B; p < 0.01), but had no effect on the lung resistance (R) (Fig. 2C). To study the effects of argon under conditions of increased PVR and to mimic a feature of PH [38][39][40] , ET-1 was added to the perfusion buffer (final concentration: 20 nM) as soon as baseline values were stable for 10 minutes. Again 10 minutes later, argon-ventilation was started. Aside from the contractile effects of ET-1 on the pulmonary vascular bed, ET-1 provokes bronchoconstriction 38 .
ET-1 significantly increased P PA , PVR, P cap , R pre and R post ( In addition, ventilation with argon did not alter the broncho-pulmonary effects of ET-1 on TV, C or R ( Fig. 2A-C).

The effects of argon on ET-1-induced lung oedema and vascular permeability.
Argon affected the formation of ET-1-induced lung oedema (Fig. 1F). To distinguish if this effect derives from reduced P cap (Fig. 1C) or from reduced vascular permeability, we perfused some lungs in addition and determined the filtration coefficient (K fc ). Perfusion with ET-1 increased K fc (p < 0.01; Fig. 3) and this effect was highly attenuated, if lungs were ventilated with an argon-mixture (p < 0.01, Fig. 3). Next, we studied in PAs, if activation of GABA-receptors plays a role within argon-induced relaxation. Thus, PCLS were treated with the GABA A -receptor inhibitor gabazine or the GABA B -receptor inhibitor saclofen prior to exposure to argon. Control PCLS only underwent argon-gassing. Regarding IVA, the gabazine/argon group did not differ from the argon group (Fig. 4D), although a trend towards gabazine-induced inhibition of argon-induced relaxation was observable. In contrast, GABA B -inhibition interacted with the relaxant effect of argon ( Fig. 4E; p < 0.01) and PAs even contracted slightly.  (Fig. 5A). We wanted to highlight if activation of GABA-receptors contributes to the effect of argon within ET-1-induced contraction of PAs, thus, we pre-treated PAs with gabazine and ET-1 (Fig. 5D) or with saclofen and ET-1 (Fig. 5E) prior to their exposure to argon. Inhibition of GABA A -receptors (gabazine) prevented the effect of argon on ET-1-induced contraction of PAs ( Fig. 5D; p < 0.05) and even increased it ( Fig. 5D; p < 0.01). Inhibition of GABA B -receptors (saclofen) also prevented the effect of argon on ET-1-induced contraction ( Fig. 5E; p < 0.05), but did not increase it ( Fig. 5E; p > 0.05). In contrast, if PCLS were not exposed to argon, inhibition of GABA A/B -receptors did not affect the tone of naïve PAs (data not shown) and did not alter the contractile effect of ET-1 in PAs (Fig. 5F).

Modulation of ET-1-induced contraction in PCLS:
Role of GABA-receptor activation. Inhibition of GABA A/B -receptors did not modulate the contractile effect of ET-1 in PAs (Fig. 5F). Next, we studied, if activation of GABA A/B -receptors alters ET-1-induced contraction in PAs, PVs or airways. Thus, PCLS were treated with ET-1 alone, ET-1/gabazine or ET-1/saclofen prior to the treatment with the GABA A -receptor agonist muscimol or the GABA B -receptor agonist baclofen (R/S baclofen). In PAs, muscimol decreased the contractile effect of ET-1 compared to their exposure only to ET-1 ( Fig. 6A; p < 0.001). The effect of muscimol was prevented, if GABA A -receptors were blocked by gabazine ( Fig. 6A; p < 0.001). Accordingly, exposure to the GABA B -receptor This effect was prevented, if GABA B -receptors were blocked by saclofen ( Fig. 6B; p < 0.001). In PVs, muscimol did not attenuate ET-1-induced contraction ( Fig. 6C; p > 0.05), but baclofen reduced it ( Fig. 6D; p < 0.001) which was also prevented by saclofen (Fig. 6D). In addition, muscimol or baclofen did not alter ET-1-induced bronchoconstriction ( Fig. 6E/F).  PAs were gassed with argon and analysed subsequently. Argon did not change the tone of naïve human PAs (Fig. 7A). ET-1 (100 nM) contracted PAs to 23% of IVA ( Fig. 7B; p < 0.001) and simultaneous exposure to argon did not alter the initial contractile effect of ET-1 (Fig. 6B). However, if PAs were exposed for 24 h to argon, the contractile effect of ET-1 was attenuated ( Fig. 7B; p < 0.05).

Discussion
In this study, argon relaxed the pulmonary circulation. Gassing with an argon-mixture (argon 74%, CO 2 5%, O 2 21%) reduced the tone of rat PAs and lowered ET-1-induced contraction in PAs from rats or humans. In IPLs, ventilation with an argon-mixture reduced the ET-1-induced increase of P cap , R post and the W/D-ratio. Regarding argon-induced relaxation, GABA-receptors appear to be involved, as 1) inhibition of GABA B -receptors prevented the relaxant effect of argon in naïve PAs and 2) inhibition of GABA A/B -receptors blocked the attenuating effect of argon on ET-1-induced contraction. Beyond that, GABA A/B seems to interact with ET-1, as stimulation of GABA A/B -or GABA B -receptors reduced ET-1-induced contraction in rat PAs or PVs, respectively. In the IPL, argon exerted some effects on the airway tone which were not confirmed in PCLS and discussed later. In rat PAs, argon exerted relaxation and reduced the contractile effect of ET-1. Both effects were evident, if argon-gassing was performed for 2 hours, whereas argon did not alter the tone of PVs, emphasizing the different response of PAs or PVs to various stimuli 30,32,41 . In line with our results from rats, argon attenuated the contractile effect of ET-1 in human PAs, although a longer duration of argon-gassing was necessary, but it did not relax naïve human PAs. The differential behaviour of PAs from both species might be due to several reasons.
(1) ET-1-induced contraction differed among PAs of rats or humans and was about 63% or 25% of IVA, respectively. This fact could explain the delayed effect of argon on ET-1-induced contraction in human PAs. (2) Rat PAs relaxed without pre-contraction suggesting a certain resting tone, as it was shown for PVs from guinea pigs 30,31 .
(3) Human PAs only relaxed due to argon, if they were pre-contracted, as it was shown for milrinone 31 and other relaxant stimuli (our own unpublished data). Most likely, they do not dispose of a resting tone, 4) PAs from both species belong to different parts of the pulmonary arterial bed. PAs from rats derive from a central part of the lung, whereas human PAs derive from a more peripheral part of the lung. Regarding the diverging responses to argon, we can expect species-dependent differences 31,36 , but also that various pulmonary vascular segments react differently to similar stimuli 42 .
Despite the intriguing effects of argon on the pulmonary arterial tone of rats or humans, argon did not influence P PA or R pre , regardless if PVR was increased by ET-1 or not. In the IPL, P PA represents the central part of the  pulmonary arterial bed which corresponds to rat PAs studied in PCLS, whereas R pre rather displays peripheral PAs. So, it was somewhat unexpected that ventilation with argon did not affect P PA in untreated lungs or did not lower the effect of ET-1 on P PA . Regarding these discrepancies, the following ideas should be considered. When PCLS are exposed to argon the argon-mixture reaches the PAs, PVs or AWs directly, thus allowing the study of argon's effects with confidence. Unlike PCLS, inhaled argon initially must pass the pulmonary vasculature at the capillary and postcapillary level where, argon effectively relaxed the tissue targets. However, it is uncertain, if argon acts in other parts of the pulmonary circulation which are more distant to the alveolo-capillary membrane to the same extent, or whether higher concentrations are required, e.g. realised by an extended application period. Potentially, the pulmonary arterial relaxant effects of argon shown in PCLS are less relevant for clinical practice. Irrespective of these considerations, argon did not deteriorate pulmonary haemodynamics, as it was shown for xenon 43,44 . It rather exerted beneficial effects on P cap , R post and oedema formation which are not debatable and of interest in patients with LHD often suffering from postcapillary PH and/or lung oedema 45 .
In the IPL, argon reduced the ET-1-induced increase of P cap and prevented the effects of ET-1 on the W/D-ratio and on K fc . Apart from that, argon reduced the ET-1-induced increase of R post resembling the postcapillary vascular bed, and thus the smallest PVs. Our data from PCLS (rats) do not reflect these results, as argon did not relax PVs and did not attenuate the contractile effect of ET-1 in PVs. Nonetheless, they become more distinct, if some other factors are considered. (1) R post is determined by the tone of the smallest PVs which are not reflected in PCLS, as PCLS from rats allow the study of central, large PVs, whereas human PCLS enable the study of more peripheral, but not the smallest PVs. (2) P LA reflects the central pulmonary venous system. However, if constant flow is applied during negative pressure ventilation, it is essential to establish a pressure balancing chamber in the perfusion outflow to prevent negative pressure lung oedema. This pressure balancing chamber is connected by tubes to the artificial thorax chamber. Hence, P LA conforms to the pressure in the thorax chamber. Finally, we studied in PCLS central or medium-sized PVs, whereas in the IPL, the smallest parts of the pulmonary venous bed were addressed.
Here, argon lowered the ET-induced increase of R post , but did not influence PVR. At first look this appears to be a discrepancy, but it is explainable. PVR is determined as followed: PVR = (P PA − P LA ) × 80/flow. Due to the facts that (1) argon had no effect on P PA , (2) P LA was fixed due to the pressure balancing chamber and (3) the flow was constant, PVR could not change. In contrast, R post is calculated as followed: R post = (P cap − P LA )/flow. Based on the facts that argon lowered the ET-1-induced increase of P cap , whereas P LA and the flow remained constant, R post unavoidably decreased. We hypothesise that argon would have decreased the effect of ET-1 on PVR, if we had applied positive pressure ventilation.
Other studies addressing the pulmonary vascular effects of argon are scarce. Martens et al. 27 studied the organoprotective effects of argon in a porcine model of ex vivo lung perfusion. PVR was increased by a warm ischaemic period of 2 hours. Afterwards, argon ventilation was started, but without effects on PVR. In a similar work, Martens et al. 28  In this study, PVR was increased by ET-1 to mimic a characteristic of PH. Notably, increased ET-1-levels play a role in sepsis, sepsis-related organ dysfunction [46][47][48] and within the pathogenesis of acute lung injury 46,49 . This issue is emphasised by the fact that ET-1-antagonists attenuate the occurrence of acute lung injury [50][51][52] . Here, argon ventilation completely prevented the effect of ET-1 on the W/D-ratio and on K fc . Thus, argon prevented the formation of lung oedema by a decrease of P cap which is the main pressure driving fluid from the pulmonary capillaries to the interstitium 53 and by decreased vascular permeability. This topic is clinical relevant, as patients suffering to neurological illness often develop neurogenic lung oedema being crucial for their prognosis 54 . Our Possibly, the diverging results rely on the various modes of induction of lung oedema. These conflicting results suggest that argon should be further explored within acute lung injury or lung oedema. Argon-induced downstream signalling is unexplored. In view of neuroprotection, a role of ERK1/2, PI3K-AKT 3,19,20 , TLR2/4 17,21,22 or Bcl-2 18 is discussed, though all these targets are fairly unspecific. The anaesthetic effect of argon is referred to the activation of GABA A -receptors 24 . GABA (γ-aminobutyric acid) is the main inhibitory neurotransmitter in the mammalian brain 55 . Beyond that, GABA-receptors are found in the lungs 56,57 . There, activation of GABA A -receptors leads to airway smooth muscle relaxation [58][59][60][61][62] and plays a role in the fetal development of the lung 63,64 .
Here, inhibition of GABA A/B -receptors did not affect the tone of naïve PAs and did not alter the contractile effect of ET-1 in PAs. Thus, the basal activation of GABA A/B -receptors appears to be not relevant. However, inhibition of GABA B -receptors (saclofen) reduced the relaxant effect of argon in rat PAs indicating a certain role of GABA B -receptors within argon-induced relaxation. The relevance of GABA for the regulation of the pulmonary vascular tone is supported by Starke et al. 65 who proved in PAs from rabbits that their contractile force is reduced, if PAs were treated with GABA and further, that GABA A -inhibition by bicuculline or picrotoxin did not prevent this effect. Hence, a dominant role of GABA B appears to be possible. Here, argon lowered the contractile effect of ET-1 in rat PAs and this effect was prevented, if GABA A/B -receptors were blocked by gabazine or saclofen. These data allow us to conclude that argon relaxes rat PAs via activation of GABA A -and GABA B -receptors. They are in line with data from Kaye et al. 66 , who found in the feline pulmonary vascular bed that both the GABA A -agonist muscimol and the GABA B -agonist SKF-97541 relaxed the pulmonary vascular bed, if it was pre-contracted with the thromboxane analogue U46619. Regarding the pulmonary vasorelaxant effects of GABA, further studies are lacking. Though, Suzuki et al. 67 showed that monocrotaline-induced pulmonary vascular remodelling was attenuated, if rats were pre-treated with GABA leading to decreased levels of norepinephrine. To the best of our knowledge, the presence of GABA-receptors has been proven in the lung 56,57 , but not specifically in the pulmonary circulation, even if the activity of GABA-transaminase was verified in PAs or PVs of guinea pigs, with dominance for PAs suggesting the presence of GABA-receptors 68 . In addition, there is evidence that stimulation of GABA-receptors alters the tone of systemic vessels 69,70 .
Beyond the pulmonary vasorelaxant effects of argon, argon reduced the effects of ET-1 on the formation of lung oedema and on K fc . Within this context, the role of GABA is supported by several studies indicating its protective effect on the development of lung oedema, e.g. Chintagari et al. 71 reported that intratracheal instillation of GABA attenuated the effect of high-tidal volume ventilation on the formation of lung oedema by an increased alveolar fluid clearance. Conversely, this effect was prevented if GABA was instilled together with the GABA A -antagonist bicuculline 71 . However, there is also evidence that activation of GABA A -receptors aggravate lung oedema 64 . These contrasting results might be explained by a switch of the Cl − conductance pattern of GABA A -receptors according to the intracellular Cl − concentration 63 . In addition, Zhang and colleagues 72 showed that propofol also acting on GABA A -receptors 73 reduces the occurrence of neurogenic pulmonary oedema.
In the IPL, we found a significant reduction in TV and C with the administration of the argon-mixture. These changes should not be related to the different viscosity of both gases, as (1) we performed a viscosity based correction of our data (explained in the method section) and (2), if we did not correct the data, TV and C would have been rather increased due to the higher viscosity of the argon-mixture. Our results from PCLS do not show any changes of the airway tone due to the exposure to the argon-mixture. They are further in line with others 12,27-29 who did not find altered lung mechanics or blood gas analyses due to inhalation of argon. Hence, it must be questioned which phenomena might account for the noticed reduction of TV and C?
The ventilation of the IPL is initiated by a negative pressure in the lung chamber. This amounts to a pressure controlled ventilation mode, such that the TV normally should be the same (based on the lung compliance) independent of the viscosity of the argon-mixture in spite of the change in airway resistance, though the time to filling could increase. However, due to the physiologically realistic breathing frequency of 70 per minute, the increased time to fill might have caused the small difference in TV that was observed.
Though, regarding the role of GABA A/B -receptors within the pulmonary vasorelaxant effect of argon, it is somewhat unexpected that argon did not exert bronchorelaxation, as activation of GABA A -receptors was shown to relax airway smooth muscle in several studies [58][59][60][61][62] . In view of the pulmonary circulation, it seems that argon-mediated activation of GABA B -receptors is more dominant than stimulation of GABA A -receptors. Anyhow, if argon stimulates GABA A -receptors in PAs, we assume that this should be also the case in AWs. Possibly, the lack of bronchorelaxant effects of argon relies on the intensity of ET-1-induced bronchoconstriction. Obviously, ET-1 contracted PAs and PVs to 60-65% of IVA, whereas the AWs contracted to 10-15% of IAA. Most probably, bronchoconstriction was too strong for a later relaxation.
In rat PAs, activation of GABA A/B -receptors (muscimol/baclofen) reduced the contractile effect of ET-1. In contrast, in PVs this was only the case, if GABA B -receptors were stimulated. Conversely, muscimol and baclofen did not alter ET-1-induced contraction, if GABA-receptors were blocked with gabazine or saclofen, emphasising the specific activation of GABA A/B . In this view, it must be questioned if ET-1 acts at all on GABA A/B -receptors, or rather if activation of GABA A/B -receptors alters ET-1-induced contraction. From the literature there is some evidence that GABA interacts anyhow with ET-1, e.g. it was reported 74 that the application of the GABA A -antagonist bicuculline led to the generation of lung oedema. In that study, ET-1 levels were increased in the bronchoalveolar lavage of bicuculline-treated rats 74 and conversely, the occurrence of lung oedema was attenuated by phosphoramidon or by the ET A -antagonist BQ-123 74 . In addition, GABA reduces the release of norepinephrine 65,67 which highly contributes to the formation of neurogenic lung oedema 75,76 .
In conclusion, argon decreased the pulmonary vascular tone of the rat, if PVR was enhanced, but it did not affect the airway tone. In view of the pulmonary vasorelaxant potential of argon, activation of GABA A/B -receptors plays a pivotal role. Finally, our results support the application of argon for neuroprotection in patients with Scientific RepoRts | (2019) 9:1902 | https://doi.org/10.1038/s41598-018-38267-y critical pulmonary haemodynamics based on PH, RV failure or LHD. The relevance of our findings is strengthened by the fact that argon also relaxed human PAs.

Methods
Animals and human lung tissue. Female Wistar rats (250 ± 50 g) were purchased from Charles River (Sulzfeld, Germany) and used as lung donors. Rat lungs were randomly assigned to one of the groups. Human PCLS were prepared from patients undergoing lobectomy due to cancer. After pathological inspection, cancer free tissue from a peripheral pulmonary part was used. None of the patients showed any signs of PH (echocardiographic or histological evaluation). The study was approved by the local ethics committee (EK 61/09) of the Medical Faculty Aachen, Rhenish-Westphalian Technical University Aachen. All patients gave written informed consent.
All animal studies and experimental procedures were approved by the Landesamt für Natur, Umwelt und Verbraucherschutz North Rhine-Westphalia (ID: 84- IPL: Preparation. IPLs were prepared as described before 31,35,77 . Briefly, female rats were anaesthetised with 95 mgkg −1 pentobarbital (Narcoren; Garbsen, Germany) and bled, if reflex checks were unresponsive. The trachea was cannulated and the lungs were ventilated with positive pressure (70 breath/min). In addition, a PEEP of 3 cmH 2 O and an I:E of 1:1 were applied. As soon, as cannulas were inserted into the pulmonary artery (inflow) and the left atrium (outflow), the lungs were perfused at constant flow (20 ml/min) with 200 ml Krebs-Henseleit buffer, containing 2% bovine serum albumin, 0.1% glucose, 0.3% HEPES and 50 nM salbutamol to prevent bronchoconstriction 78 . The buffers' temperature was kept at 37 °C by a water bath and the pH was maintained between 7.35 and 7.45 by gassing with CO 2 . After pulmonary ventilation and perfusion was established, heart and lungs were removed and set into a negative-pressure chamber which was adjusted for P min = −7 cmH 2 O and P max = −2 cmH 2 O. To prevent lung oedema during constant flow perfusion and negative pressure ventilation, a pressure balancing chamber was set in the perfusion outflow and connected by tubes to the artificial thorax chamber. To prevent atelectasis, every 5 minutes a deep breath was applied. Airway (TV, C or R) and pulmonary haemodynamic parameters (P PA , P LA and flow) were recorded with the Pulmodyn Software 2.0 (Hugo-Sachs Elektronik Harvard Apparatus, Germany). If all parameters were stable, the PVR was enhanced with ET-1 (final buffer concentration: 20 nM) 79 .

IPL: Calculation of airway parameters.
The pressure inside the thorax chamber was measured by a pressure transducer (Hugo-Sachs Elektronik Harvard Apparatus, Germany). The inhalation flow (Q) was measured by a pneumatograph (Hugo-Sachs Elektronik Harvard Apparatus, Germany). Finally, TV was calculated from the integration of Q. The compliance (C) of the lung expresses its elasticity. C is defined by the TV and the change (Δ) of the transpulmonary pressure (P tp ). In the IPL, ΔP tp reflects the difference of the maximal and minimal pressure (P) in the thorax chamber. So, C was calculated: C = TV/(P max − P min ) and R (resistance) was calculated by the inhalation flow (Q) in relation to P max and P min : R = (P max − P min )/Q 35 .

IPL: Ventilation with argon and correction of airway parameters.
To ventilate the lungs with argon, a pressure regulator and a flow meter were used. Argon (flow 0.3 L/min) was applied via a tube which was connected to the pneumatograph. Inhalation flow was measured using a pneumatograph consisting of a single tube with small enough diameter to maintain laminar flow such that the pressure drop across the tube is linearly proportional for the flow rate. The complication in this application occurs because the proportionality is dependent on the gas viscosity (µ); i.e., if the gas viscosity is increased the pressure drop needed to drive the flow would be increased. The pneumatograph was originally calibrated for air; thus for the argon mixture that has a greater viscosity (2.227 kg/(s-m)10 −5 for argon at 32 °C versus 1.84 kg/(s-m)10 −5 for air at 32 °C) the measured pressure drop will be greater than that for air at the same flow rate. Therefore, a recalibration of the pneumatograph for the argon-mixture would be necessary; however this is not applicable without interrupting the experiment. Another possibility is to correct the pressure drop data by multiplying by the viscosity ratio of air over argon to obtain the flow rate. Alternatively, TV, R and C can be corrected. We corrected TV, R and C as followed: TV corr = TV meas × (µ air /µ argon-mix ); R corr = R meas × ( µ air /µ argon-mix ) and C corr = TV corr /(P max − P min ) 80 .
IPL: Calculation of PVR, precapillary and postcapillary resistance. PVR was calculated as followed: IPL: Wet-to-dry ratio (W/D-ratio). After IPLs were perfused for 2 h, the wet weight of the right superior lobe was recorded and subjected to drying at 60 °C for 72 h. The dry weights were monitored and the W/D-ratio was calculated.

IPL: Assessment of the vascular permeability by determination of the filtration coefficient.
To distinguish, if lung oedema derives from increased P cap or increased vascular permeability, the capillary filtration coefficient (K fc ) was determined as described in reference 35 . Measurements were performed at 0 and 120 minutes of the perfusion using the following equation: K fc = (dweight/dtime)/dP cap . Due to the fact, that weight gain measurements do not allow the simultaneous application of the double occlusion method, P cap was calculated according to the Gaar equation 82 : P cap = P LA + 0.44 (P PA − P LA ).
Scientific RepoRts | (2019) 9:1902 | https://doi.org/10.1038/s41598-018-38267-y PCLS of rats and humans: Preparation. Rats received intraperitoneal anaesthesia with pentobarbital, which was verified by missing reflexes. Thereafter, they were prepared as described before 31,36 . Rat lungs were filled via the trachea and human lungs were filled via a main or lobar bronchus, respectively, with 1.5% low-melting agarose. Afterwards, they were cooled on ice. Tissue cores (diameter 11 mm) were prepared and cut into about 250 µm thick slices with a Krumdieck tissue slicer (Alabama Research & Development, Munford, USA). PCLS were incubated over night at 37 °C and repeated medium changes were performed to wash out the agarose.

PCLS: Treatment and videomicroscopy.
To study the role of GABA within argon-induced relaxation, PCLS were treated with the GABA A -receptor antagonist gabazine (5 µM) 83 or with the GABA B -receptor antagonist saclofen (5 µM) 84 (Fig. 4D/E). To study the relaxant effect of argon in pre-contracted PAs, PVs and AWs, rat PCLS were pre-contracted with 200 nM ET-1 (Fig. 5A-C) and human PCLS were pre-contracted with 100 nM ET-1 (Fig. 7B). To study the role of GABA-receptors within argon-induced relaxation in ET-1 pre-contracted PAs, PCLS were simultaneously pre-treated with 200 nM ET-1 and 5 µM gabazine (Fig. 5D) or with 200 nM ET-1 and 5 µM saclofen (Fig. 5E). To study the effect of GABA-receptor inhibition or activation within ET-1 induced vasoconstriction and bronchoconstriction, PCLS were pre-treated with ET-1 and gabazine (Fig. 6A), ET-1 and saclofen (Fig. 6B/D) or ET-1 alone (Fig. 6C/E/F) prior to the treatment with the GABA A -receptor agonist muscimol (5 nM) 85,86 or the GABA B -receptor agonist baclofen (5 µM) 87 . In PCLS, all changes of the IVA and IAA were quantified in % and indicated as "Change of IVA [%]" or "Change of IAA [%]". Thus, an IVA < 100% indicates contraction and an IVA > 100% indicates relaxation. In order to compare the effect of argon in pre-treated vessels, the intraluminal area was defined after pre-treatment again as 100%. In the graphs, all pre-treatments were indicated. The intraluminal area of PAs, PVs and airways was monitored with a digital video camera (Leica Viscam 1280, Leica DFC 280). The images were analysed with Optimas 6.5 (Media Cybernetics, Bothell, WA).
PCLS from rats and humans: Argon gassing. PCLS were transferred into 24 well plates (1 ml medium per well). In order to treat PCLS with argon (Argon 74%, CO 2 5%, O 2 21%) or air mix (N 2 74%, CO 2 5%, O 2 21%), we used the modular incubator chamber (Billups-Rothenberg, USA) with a filling volume of 6 liters. To avoid evaporation, we added 20 ml of purified water in a petri dish inside the chamber. Next, the incubator chamber with the PCLS inside was purged with the appropriate gas mixture and a gas flow rate of 20 L/min for 5 minutes; the outlet and inlet ports where closed with plastic clamps. Alkalisation of the incubation medium was prevented, as all gas mixtures were enriched with 5% CO 2 . Afterwards, the whole chamber was transferred in a heat CO 2 incubator. In order to investigate the effects of argon on pre-constricted airways or vessels, we pre-treated PCLS with ET-1 and made images at 1 h, 2 h, 3 h, 6 h and even 24 h, in case of human PCLS. To study the role of GABA within the relaxant effect of argon, we pretreated PCLS with the GABA inhibitors gabazine or saclofen and made images according to the pre-treatment with ET-1. Images were recorded by videomicroscopy and the IVA/IAA was calculated with Optimas 6.5.
Reagents. ET-1 was purchased from Biotrends (Wangen, Switzerland). The potency of ET-1 differs strongly with age and lot numbers. GABA receptor agonists/antagonists and standard laboratory chemicals were from Sigma-Aldrich (Steinheim, Germany). Gas mixtures were delivered from Air Liquide GmbH (Simmerath, Germany) or Linde Gas AG (Pullach, Germany). statistical analysis. Statistics was conducted using SAS 9.2 (SAS Institute, Cary, North Carolina, USA) and GraphPad Prism 5.01 (GraphPad, La Jolla, USA). All data were analysed by a linear mixed model analysis, except Figs 1F and 3 which were analysed by the Mann Whitney U Test. P-values were adjusted for multiple comparisons (false discovery rate) and presented as mean ± SEM. N indicates the number of animals or lung lobes. P < 0.05 are considered as significant: *p < 0.05, **p < 0.01, ***for p < 0.001.

Data Availability
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.