Sign in

Possible Mechanisms of Cryoprotective Effect of Xenon

To begin let me thank the organizers of this great event—Danke schön!

This is the first of a series of three reports from our group’s research. So I'll provide a very short preface to the whole theme. [Slide 2 - Abstract] Our common topic is use of xenon (and, possibly, other inert gases) in cryobiology. The use of xenon has been first proposed - at least as far as we know – by American chemist Robert Prehoda in his book 'Suspended animation...' as early as in 1969.  [Slide 3 - Prehoda] He motivated this proposal by suggestion that in presence of xenon water would form clathrate crystals of xenon hydrate rather than that of ice without discussion on why the first ones would do less harm than the second. Since that there have been scattered reports on xenon's cryoprotective abilities. Some of them are listed in the “References” slide at the end of presentation while others were private communications on studies that haven't been published because of their inconclusive outcomes. Among them the works of Vl. Kovanov's group in 1980's should be mentioned. [Slide 4] They claimed restoration of heartbeat in a rat after submerging it in liquid nitrogen for 3 hours and received a patent on the method. As far as we know, nobody has ever reported a confirmation of these results.

Xenon is a somewhat a mystical substance. Being all but absolutely chemically inert it displays some prominent physiological impacts. [Slide 5] The best known of them and routinely used in clinics in Germany and Russia is it's anaesthetic action. In fact it looks to be the best inhalational anaesthetic currently known, with the only shortcoming being its relatively high cost. This ability is known for more than a hundred years, but the underlying mechanism remains under discussion. Surely, there are certain hypotheses, but none of them are generally accepted by now.

These properties look especially  attractive when compared to common cryoprotectants. [Slide 6]  In case xenon really has reasonable cryoprotective abilities, then all the issues in this list can be resolved, including the worst of them, which is toxicity.

In our studies we tried to answer three questions related to cryobiology [Slide 7]:

  1. Does xenon really act as a cryoprotectant?
  2. If 'Yes', then what are the mechanisms of such an action?
  3. How can it be applied in practice, especially having in view the cryopreservation of bulk biological objects.

To answer these questions we initiated two lines of studies - a theoretical one and an experimental one. Both are currently in their initial states, so our results are somewhat preliminary, but we already see some exciting new results. This particular report is dedicated to molecular dynamics study of both (xenon + liquid water) and (xenon + liquid water + ice) systems at temperatures about and below freezing point.

There are several (mutually nonexclusive) possible mechanisms of cryoprotective ability of xenon [Slide 9]:

  1. Accumulation of xenon in cell membranes, preventing liquid crystal to gel phase transition and thus maintaining membrane elasticity or/and in protein molecules' hydrophobic "pockets" thus preventing them from denaturation.
  2. Formation of xenon hydrate crystals which compete with ice ones for water molecules and are for some reason less destructive for cells and tissues. For example they can be finer and/or have less damaging shape, e.g. granular instead of needle-like that of ice.
  3. Ice-blocking effect of xenon due to its accumulation in the ice-water interface zone.

Vitrification of xenon-water solution. Its conditions may differ from that of pristine water. As for mechanism 1, it is shown in paper [Slide 10] by Booker and Sum that xenon atoms preferentially localize in the hydrophobic core of the lipid bilayer, inducing substantial increases in the area per lipid and bilayer thickness. Xenon depresses the membrane gel–liquid crystalline phase transition temperature, increasing membrane fluidity, so at least that mechanism can be considered as confirmed.

In our study we have concentrated on mechanisms 2, 3 and 4. Our results have been published in the Journal of Chemical Physics [Slide 8], so here I provide only a concise summary of our methods; one can find a detailed description of them in the paper. Instead I'll concentrate on aspects of our study, related to cryobiology rather than chemistry or physics, some of which are not covered in the paper.

Let us begin from the xenon-water system. [Slide 11] We have simolated both normal (280°K, 275°K) and supercooled (270°K, 265°K, 260°K, 255°K) supersaturated (containing 1, 2, 3, 4, and 5 %% of xenon) solution. Such high concentrations may seem unrealistic; the fact is that we can't say exactly what pressures are necessary to produce such concentrations since we couldn't find data of xenon solubility in supercooled water [Slide 12]. But if a known semi-empiric dependency extends into the area of supercooled water, than we can estimate necessary pressures as that in range of tens to hundreds of bars, so they are absolutely technically feasible.

Now [Slide 13] let us return to the table which displays investigated combinations of temperatures and concentrations and time spans for each of these runs. The less the temperature, the more the time necessary for the parameters of the system under investigation to quasi-equilibrate to a metastable state. At 250°K all the processes become so slow that we couldn't achieve any metastable-looking state even after spending several hours of supercomputer time.

In all cases we see formation of hydration shells around atoms of xenon. Typical shell consists of 20 to 24 molecules of water with an average of ~21.5. [Slide 14 RDF] [Slide 15 Paschek] These shells are not stable in the sense that water molecules come and go, but when water molecule goes, its place in the shell is promptly filled with another water molecule.

In systems with xenon concentrations greater than 3% we can see homogeneous nucleation of xenon crystal hydrate [Slide 16] (corresponding runs are accentuated by green color) [Video]. We can observe two "failed attempts" to nucleate a crystal followed by two successful nucleations. On [Slide 16] we can see the final result – two  microcrystals of xenon clathrate hydrate.

Here we can make several important observations:

  1. In all cases where crystal hydrate of Xenon has been observed it's crystals formed almost immediately, on timescale of nanoseconds, whereas in none of our runs have we observed formation of ice crystals.
  2. In some cases more than one crystal formed in such a small volume as our computational cage.
  3. Soon after formation of the crystal it's growth halted or at least greatly slowed - presumably due to the (not so big) decrease of xenon concentration in the liquid phase.
  4. Xenon hydrate microcrystals look to be of granular form.

So we can return to the list of possible mechanisms of cryoprotection [Slide] and mark the first of them as probable.

To investigate xenon's ability to act as an ice blocker we have simulated Liquid water + Xenon + Ice system.  [Slide]

In the presence of xenon equilibrium water/ice temperature diminishes for approximately 1K per 1% of xenon which is in accordance with cryoscopic law. Atoms of xenon can penetrate into ice for as far as 2 crystal cells, but they are always expelled from it in the process of crystal growth. Liquid/crystal interface is more loose in presence of xenon and we can see that xenon really concentrates in the interface zone with its concentration becoming up to twice as big as in the bulk liquid. But we couldn't observe any impact of xenon presence on the rate of crystal growth or melting, so [Slide] we have to mark the second mechanism as "unlikely".

To evaluate the possibility of mechanism #3 – vitrification – we have to return to homogeneous system, containing neither ice nor xenon hydrate crystals. In vitreous state all the chemical processes come to a halt because diffusion of molecules stops. So it is necessary to determine diffusion constants. On the slide [Slide] you can see the temperature dependency of diffusion constant of xenon for various concentrations. Graphs for 4% and 5% are of little interest since at these concentrations we observe formation of hydrate crystals – they are represented by kinks on these graphs. In the 1%, 2% and 3% systems we can see a dramatic diminishing of diffusion constant with cooling. It falls from 280K to 255K to less than 1/10. At 250K all processes are so slow that it was impossible to reach a quasi-equilibration in the affordable simulation time.

We extrapolated diffusion constant from three ‘coolest’ points on the graph – those of 265, 260 and 255K. [Slide] The resulting parabola hits zero at ~249K, which can be interpreted as the temperature of sol/gel phase transition, that is vitrification. At that point diffusion comes to a halt, extrapolated viscosity rises to infinity and all chemical processes stop.

Of course this is just an extrapolation of a simulation and we can’t be sure of it before it is confirmed experimentally. And it may occur not so easy to cool a bulk biological object such as organ for transplantation or whole cryopatient down to ~-25C avoiding crystallization in the process. One obvious approach may be in combination of xenon with reasonable concentrations of more traditional cryoprotectants. The latter would inhibit both ice and hydrate crystals formation allowing vitrification at technically feasible rates of cooling. Another approach is application of extracellular cryoprotectants to partially dehydrate the cells with the same effect. And of course, these approaches can be combined.

There is a great area for future investigations and we are absolutely open to any form of collaboration – from funding to coordinated studies, both theoretical and experimental.