Argon: Systematic Review on Neuro- and Organoprotective Properties of an “Inert” Gas

Argon belongs to the group of noble gases, which are regarded as chemically inert. Astonishingly some of these gases exert biological properties and during the last decades more and more reports demonstrated neuroprotective and organoprotective effects. Recent studies predominately use in vivo or in vitro models for ischemic pathologies to investigate the effect of argon treatment. Promising data has been published concerning pathologies like cerebral ischemia, traumatic brain injury and hypoxic ischemic encephalopathy. However, models applied and administration of the therapeutic gas vary. Here we provide a systematic review to summarize the available data on argon’s neuro- and organoprotective effects and discuss its possible mechanism of action. We aim to provide a summary to allow further studies with a more homogeneous setting to investigate possible clinical applications of argon.


Introduction
Argon belongs to the noble gases and generally is regarded as an inert, non-reactive element. Even its name (from the Greek "αργός"-inert) refers to its chemical inactivity. In fact, biological effects of the noble gases including argon have been identified starting in the 1930s: its narcotic properties under hyperbaric circumstances were described beginning with studies investigating argon as a possible breathing gas for divers [1]. Recently, neuroprotective and organoprotective features have been identified [2][3][4][5][6].
In general, most promising therapeutics-especially neuroprotectants-identified through preclinical studies have failed to demonstrate efficacy in clinical trials due to heterogeneous experimental settings, inadequate sample sizes, inappropriate time and dosage of application and so on [7,8]. Concerning argon's beneficial properties, most of the evidence has been accomplished by in vitro, in vivo and rarely human studies. Again, the multitude of anecdotal reports and experimental models applied hinders the overall assessment of argon's therapeutic potential but also its possible side effects. Therefore we performed a systematic review on the current literature on argon. We provide an overview of available data on argon's organoprotective and particularly its neuroprotective features as well as potential side effects. Further, we illustrate the current data on the possible mechanism of action and future perspectives for therapeutic applications of argon.

Results
The PubMed search revealed 671 hits, from which 42 records were identified as relevant for screening. The alternative databases (Embase, Scisearch, Biosys, gms) presented 1501 records using the same search strategy. Eighty-seven records were regarded relevant. Thirty-five articles had to be excluded with regard to content (review articles, comments or articles on technical applications of argon, abstracts and poster presentations); one article had to be rejected as only available in the Chinese language. Duplicates (n = 65) among the two database searches were eliminated. In Figure 1 the procedure is summarized. In total, 38 relevant full text articles were identified. Eleven out of 38 (29%) studies were conducted before, and 27 (71%) after the year 2000. Human studies are scarce (n = 6, see Table 1) and most of them had been motivated by technical considerations in the context of diving or aerospace. In vivo animal experiments dealing with the effects of argon are much more common (n = 22, summarized in Table 2) and the number of publications on in vivo data has increased recently (16 out of 22 articles have been published later than 2000). Most animal experiments were carried out with rats (16 out of 22); in two studies, Japanese quail eggs were used. In vitro studies are dominated by the use of murine organotypic brain slices (4 out of 10 studies; see Table 3). Comparisons with other noble gases were drawn in 7 (out of 22) animal studies and 5 in vitro studies. Frequently the effect of argon was compared to that of helium (n = 7) and xenon (n = 5).
In human studies, descriptions of argon's narcotic effect and the possible increase of resistance against hypoxia were most common, whereas among in vivo animal studies, the neuroprotective or organoprotective properties of argon were the main topic (11 out of 22). In general most of the studies dealt with argon's narcotic effect and the reaction of organisms to hypoxia under argon atmosphere (n = 14). Notably most of these studies were carried out before the year 2000 (9 out of 14). Neuroprotection and organoprotection are relatively new topics: All of the studies covering these topics (n = 17) were carried out after the year 2000. Neuroprotection is, with 11 articles, the field of interest most frequently highlighted in the recent years. Besides tissue protection, recent studies often dealt with the identification of argon's mechanism of action. In total, 11 investigations addressed this question with 10 of them carried out after 2000. Most studies concerning protective effects of argon and its mechanism of action were carried out using animal or in vitro models.
Notably, argon failed to show protective properties in two studies [9,10], whereas other studies on tissue protection could only demonstrate a partial benefit of argon treatment (i.e., only functional improvement or benefit under certain circumstances like timing of applications) [11,12].

Discussion
Our systematic review has highlighted various studies on argon's effects applying heterogeneous models and questions. We will discuss similarities and differences of the approaches and results.

Physiological Studies
The first descriptions of argon's biological effects arose in the context of diving medicine. Mental impairment at high pressures had been observed. Behnke and Yarbrough in 1939 tried to elucidate the role of argon in producing narcotic effects in humans [1]. The first physiological data (like respiratory resistance) could be assessed. Further studies were carried out evaluating mental performance and subjective rating of condition as measurement of narcosis with 80% argon and 20% oxygen under different pressure levels [13]. In humans, long-term effects (up to 7 days) under hyperbaric argon atmosphere were examined demonstrating improved work performance and a shift in lipid metabolism. An increased resistance to hypoxic hypoxia under argon atmosphere was suggested [14]. These results were confirmed by a study testing human oxygen consumption during physical exercise breathing a gas mix with 30% argon. An increase of oxygen consumption under argon was observed, therefore a catalytic activity of argon on oxygen kinetics was supposed [15]. Another long-term study carried out for 9 days (14% oxygen, 33% nitrogen, 54% argon and 0.2% carbon dioxide for 6 days followed by 10% oxygen, 35% nitrogen, 55% argon and 0.2% carbon dioxide) demonstrated no detrimental influence on work performance [16].
Narcotic potency was also examined in mice [9] and rats [17,18]. Therefore metabolism, oxygen consumption and resistance towards hypoxia under argon (different species and organ slices) were investigated [9,[19][20][21]. A favorable metabolic condition with a distinct energy metabolism and elevated oxygen consumption was supposed, thus resulting in increased resistance towards hypoxia [20,22]. Furthermore, an improved survival of animals under hypoxic atmosphere was demonstrated [23,20], whereas an earlier series of experiments with white mice did not indicate a beneficial effect of argon on survival [9]. Under hypoxic atmosphere containing argon a change in development (faster development and less teratogenic pathologies) was observed, which also has been attributed to the change of metabolism under argon [19,24,25].
In conclusion, argon seems to improve resistance towards hypoxia. Metabolic changes, cell membrane dependent mechanisms, recovery of mitochondrial enzymes and oxygen synergism have been discussed to explain this phenomenon. Studies concerning mechanism of action will be discussed hereafter in more detail.  [15] Ar = argon; O 2 = oxygen; N 2 = nitrogen; CO 2 = carbon dioxide; TEOAE = transitory evoked otoacoustic emission; DPOAE = distortion product otoacoustic emissions; BERA = brainstem evoked response audiometry; EcohG = electrocochleography.

Neuroprotective and Organoprotective Properties
Within a multitude of experimental models protective effects of argon were investigated: In vitro mostly fetal organotypic murine brain slices were applied. Lesion was induced either by mechanical trauma (in vitro traumatic brain injury-TBI) or by oxygen-glucose-deprivation (OGD) simulating global metabolic stress, i.e., ischemia. Mechanical as well as metabolic stress was diminished by argon application repeatedly [37,40,41,43]. The concentration of argon varied, but averaged at least 50% (in one study 50% atm was applied). Dose dependency for argon treatment after OGD was demonstrated by Loetscher and colleagues, whereas the most effective concentration after in vitro TBI was identified with 50% argon. Even delayed application of postconditioning with argon (up to 3 h after injury) still resulted in decrease of cell death compared to controls without argon treatment [41]. Without injuring the brain slices, application of 75% argon was even able to reduce cell death when compared to controls and showed a less pronounced protective effect than xenon [37]. Another organotypic model assessed hypoxic and toxic resistance of hair cells (organ of Corti) under argon containing atmosphere demonstrating an otoprotective effect [44].
In vivo the most common models are those inducing hypoxia either resulting in cerebral ischemia (by middle cerebral artery occlusion, or hypoxic ischemic brain injury with ligation of carotid artery and exposure to hypoxia) or myocardial ischemia (by LAD-left anterior descending artery-occlusion) or both (by cardiac arrest (CA) followed by delayed resuscitation (CPR)). In line with the clear protective effect after OGD-an in vitro model for cerebral ischemia-Ryang and colleagues [34] demonstrated a reduction of infarct volume and improved composite adverse outcome following argon postconditioning using an MCAO rat model. With the same model (MCAO) but different application time of argon David and colleagues [11] also showed a reduction of cortical infarct volume, but subcortical brain damage increased with argon treatment. In this connection the argon treated animals revealed worse neurological performance (compared to sham), which was found at days 1 and 2 after MCAO. This contrasts with xenon that provides both cortical (greater than argon) and subcortical neuroprotection and further showed improved neurological outcome in the same conditions of MCAO model and timing of treatment [45]. As discussed by David et al., differences between their results and those of Ryang et al. as regards to subcortical neuroprotection could be due to differences in study protocol, particularly timing of treatment (intraischemic vs. postischemic). However, in the same study of David and coworkers, protective effects of argon after OGD were confirmed and, in vivo, an attenuation of NMDA-induced brain damage was shown. Neuroprotective properties after hypoxia (hypoxic ischemic brain injury rat model) were confirmed by Zhuang and colleagues [35]. A more pronounced beneficial effect regarding cell viability for postconditioning with argon was described vis-a-vis nitrogen and even xenon. Combining some features of the aforementioned models, some groups use a resuscitation model to induce cerebral ischemia: In pigs and rats postconditioning with 70% of argon resulted in improved neurological outcome [12,28]. The morphological extent of brain damage (at least for some regions) was reduced compared to controls. In rats dose dependency of the beneficial effect after resuscitation was demonstrated with better neurological outcome after treatment with 70% argon than with 40% [29].
Cardioprotective effects with decrease of infarct size were shown by an in vivo study using argon as a preconditioning drug with a rabbit model [33]. Another possible application of argon is to protect donor organs before transplantation. Rat kidneys harvested in argon saturated solution demonstrated better functional and morphological condition than controls (nitrogen saturated solution) or even xenon treated group [32].
Finally the only human study on neuroprotection was carried out to investigate the effect of argon treatment on exposure to white noise. Improved condition of the acoustic system was shown assuming an oto-and neuroprotective effect of treatment [23]. This hypothesis was strengthened by experimental data on improved hair cell survival with argon treatment.
In conclusion argon's neuroprotective and organoprotective properties were confirmed by various studies using a multitude of experimental models primarily to simulate hypoxic and less frequently mechanical cell stress. Neuroprotection is the field most commonly covered and most studies underscore the beneficial effect of argon treatment. Nevertheless, results are biased by heterogeneously applied experimental models and differences in study protocols (different timing, concentration and duration of treatment).

Mechanism of Action
Very little is known about the actual mechanism of action of argon. Abraini and colleagues investigated the involvement of GABA-receptors by examining argon's narcotic potency in rats after pretreatment with specific GABA-receptor antagonists. They discovered that in a rat model argon threshold pressure had to be increased after pretreatment with GABAA-receptor antagonists and to a lesser extent after GABAA-receptor antagonists for the benzodiazepine site. This was not the case after pretreatment with a GABAB-receptor antagonist. Thus-similar to nitrogen-involvement of GABAA-and the benzodiazepine site of GABAA-receptors, but none of GABAB-receptors, was hypothesized [26]. However, this finding is limited by the fact that Abraini and colleagues used the (hyperbaric) narcotic properties of argon as outcome parameter. Therefore it is problematic transferring the results into the area of neuroprotection, which is achieved under normobaric circumstances. Furthermore at atmospheric pressures, argon did not provoke an intracellular acidosis in macrophages that is induced by other benzodiazepine-sensitive GABAA-receptor agonists [46]. Thus, two distinct, independent methods of action are conceivable dependent on ambient pressure and response measured.
Another in vivo study correlated the extent of striatal dopamine release with the narcotic effect of argon. Decrease of striatal dopamine release was seen in parallel to gas narcosis [27]. Again, this finding relies on argon's narcotic properties not its cytoprotective properties as indicator of outcome.
Similar to xenon, which inhibits NMDA-receptors [47], this receptor type was investigated during argon treatment. Application of glycine did not reverse the beneficial effect of argon after in vitro TBI, therefore involvement of the glycine site of the NMDA-receptor in argon's mechanism of action was ruled out. Further, using electrophysiology (patch clamp technique) no effect of argon on NMDA-mediated currents was found, likewise for currents flowing through TREK-1, a two-pore-domain potassium channel [40]. In an in vivo study (resuscitation rat model), pretreatment with a KATP-channel blocker (5-Hydroxydecanoate = 5HD) also failed to impact argon's beneficial effect [29]. Therefore, neither NMDA receptors nor potassium channels seem to be involved. However, these results will have to be confirmed in further experiments.
Another in vivo study tested in a rat model of hypoxic ischemic brain injury and postconditioning with 70% argon, helium or xenon the expression of three proteins involved in the intrinsic apoptotic pathway: Bax, Bcl-2, and Bcl-xL. Treatment with argon, helium, and xenon increased the expression of Bcl-2. Surprisingly, helium and xenon, with the exception of argon, increased Bcl-xL, a prosurvival protein, whereas expression of Bax, which promotes cell death, was induced after treatment with helium [35]. Again, these results may reflect the uniqueness of each noble gas in regard to its mechanism of action. Further, noble gas modulation of prosurvival proteins has to be elucidated.
Using cultured renal tubular cells (HEK2) prosurvival proteins were investigated in vitro. After preconditioning with 75% helium, neon, argon, krypton and xenon, cell cultures were subjected to OGD. Surprisingly, only xenon treatment showed protection of cell viability. Further, prosurvival proteins (Bcl-2, pAkt -Phospho-Akt-and HIF-1α-hypoxia inducible factor 1 α) were analyzed without OGD. Expression of HIF-1 α increased after treatment with argon, while Bcl-2 and p-Akt expression were not modified. However, xenon treatment led to an increase of all the examined proteins, Bcl-2, p-Akt and HIF-1α [10]. This is in contrast to the results mentioned above and may be due to different experimental settings (in vivo vs. in vitro), different models of tissue stress (hypoxic ischemic brain injury vs. naïve cell culture) and different time points of analysis.
Multiple damaging agents were tested in an in vitro study using a human osteosarcoma cell line (U2OS). Cells were exposed to a tyrosine kinase inhibitor (staurosporine), a DNA-damaging agent (mitoxantrone) and mitochondrial toxins. Argon and xenon inhibited cell loss by staurosporine, mitoxantrone and the mitochondrial toxins, maintained mitochondrial integrity and inhibited caspase-3 expression [43]. Suppression of caspase-3 and cytochrome C once again indicates inhibition of intrinsic apoptotic pathway by argon and xenon.
Using microglial cell cultures and primary neuronal and astroglial cultures the involvement of ERK1/2 (extracellular signal-regulated kinases) with a short and enhanced activation after exposure to 50% argon was demonstrated, but no relevant influence on cytokine expression (contrary to xenon) was found [38].
Finally, protein interactions of argon have to be mentioned: Colloch'h and colleagues investigated the protein-noble gas interactions of xenon, krypton and argon [48]. Three different enzymes were studied showing gas occupancies in the order of their polarizability with highest occupancy reached by xenon and lowest by argon administration, which is similar to the results of Quillin and colleagues examining T4 lysozyme [49]. Depending on the enzyme, different mechanisms of noble gas action were demonstrated: either inhibition of the catalytic reaction through an indirect mechanism, inhibition of the catalytic reaction through a direct mechanism, or prevention of substrate binding. The considerable effects of noble gases are not completely explained by the binding through very weak non-covalent van der Waals interactions. Therefore, the authors conclude that small effects on an array of biological targets may be responsible for the biological effects of noble gases but specific effects (like neuroprotection) of the noble gases may also be due to action via one particular target, which may be specific for each noble gas [48].
In conclusion, argon may distinguish itself from xenon while possibly sharing some joint features during further signaling (like Bcl-2 involvement). Also ERK1/2-signaling plays a role in signal transduction by argon. Decidedly, argon seems not to act via NMDA-receptor signaling or via potassium channels. Although argon would act as a GABAA agonist to induce narcosis as shown in hyperbaric conditions, whether this could apply to normobaric condition as a mechanisms for neuroprotection still remains to be shown. Therefore, the precise target(s) for the biological effects generated by argon administration remains to be elucidated. Only limited evidence indicates the involvement of GABAA-receptor signaling. Finally, approaching the topic from the chemical point of view, one has to highlight the assessment of two important chemists (Nikolai Nikolajewitsch Semjonow and Cyril Norman Hinshelwood), who pointed out the oxygen-like properties of argon: the presence of argon allows reactions between phosphorous and oxygen under pressure levels, which would not happen without argon, therefore acting as sort of catalyst [50]. Thus, increase of resistance towards hypoxia may be explained by argon's oxygen-like properties as hypothesized by David and colleagues previously [11,36].

Lack of Clarity
However, while appreciating many promising details of argons possible protective actions, some discrepancies should not be overlooked: In one in vivo study under hypoxic argon atmosphere, mice did not survive longer than the control group [9]. Another in vitro study using OGD as experimental model did not disclose a beneficial effect of argon preconditioning [10]. Finally, argon treatment in rats applying MCAO resulted in one study in reduced infarct volume (including subcortical area) but in the other in increased infarct volume of subcortical area and worse neurological outcome [11,34]. During one trial the application of argon occurred within the intraischemic phase, and on another occasion after reperfusion, as David and colleagues clearly pointed out [11].
These discrepancies may be attributed to differences in the study protocol. One major problem analyzing the studies on argon is that treatment varies between pre-conditioning and post-conditioning. Even if the same "type" of treatment is applied, timing, concentration and duration of administration diverge.
Therefore, to gain more insights into argon's protective effects as well as identifying its mechanisms of action, standardizing study protocols would be advantageous. Argon's cytoprotective and special neuroprotective properties have been demonstrated in many studies. Transfer into clinics has not yet occurred due to a lack of data for argon's practical implementation and potential side effects. David and colleagues [36] tested argon in the context of tPA (tissue-type plasminogen activator) application to review a potential application in stroke therapy. Results demonstrate a dual argon effect. The somehow unexpected inhibiting effect of argon at low concentration on tPA efficiency according to the authors may be due to aforementioned interactions with proteins dependent on multiple factors like gas accessibility and affinity to hydrophobic cavities and the oxygen-like properties of argon [36]. Thus, additionally considering dual effects is necessary for further identification of the appropriate clinical administration concerning timing and duration as well as detection of the mechanism of action.

Methods
A PubMed search was carried out in June 2014 with the following search terms: neuroprotection OR organ protection OR cell death OR neuro* OR hypoxic ischemic encephalopathy OR asphyxia OR ischemia OR hypoxia OR ogd OR tbi OR protect* AND argon. Additionally, alternative databases (Embase, Scisearch, Biosys, gms) were screened for the same search terms. Afterwards duplicates were eliminated. The reference lists of review and other relevant articles were hand-searched for appropriate articles and two additional articles, which were later published online ahead of print, were included as well. Of note, Russian articles have been translated by a non-native speaker and therefore we might have caused a translation bias. Additionally, the heterogeneity of experimental settings may hinder the final appraisal.

Conclusions
Argon's neuroprotective and organoprotective properties have been demonstrated repeatedly, but still uncertainties arise from the inhomogeneity of applied models, timing and dosage of argon application.