Vitrifying a whole organ to defeat ice — and reviving it on the other side
For over half a century, cryobiology has been defined by one unsolved obstacle: ice. Cooling water-based tissue below 0°C does not gently "pause" it — it destroys it, because water expands and reorganizes into a crystal lattice with no regard for the delicate architecture of cells. Whole organs, with their intricate vasculature and centimeter-scale dimensions, have never survived conventional slow freezing intact.
When aqueous tissue is cooled below 0°C, pure water in the extracellular space nucleates into ice before the more concentrated intracellular fluid does, because dissolved solutes (salts, proteins, sugars) depress the freezing point inside cells. As extracellular ice grows, it progressively excludes solutes from the lattice, concentrating them in the shrinking pool of remaining unfrozen extracellular liquid.
This creates a steep osmotic gradient across the cell membrane: the now hyperosmotic extracellular fluid pulls water out of the cell, causing progressive cellular dehydration. Cells that shrink too far, too fast, suffer "solution effects" injury — mechanical membrane stress, protein denaturation, and dangerously concentrated intracellular electrolytes.
At typical freezer cooling rates (~1°C/min), extracellular ice crystals can grow to 50–200 micrometers — roughly 5–20 times the diameter of a single cell — turning the extracellular space into a field of razor-edged crystal blades.
If cooling continues, or if the rate is too fast for water to osmotically escape the cell, ice also nucleates inside the cell itself. Blocked by the membrane and organelles, intracellular ice forms as many small, jagged crystals that puncture membranes, rupture mitochondria and the nucleus, and shred the cytoskeleton the instant they form.
Cryobiologists describe this with Mazur's two-factor hypothesis of freezing injury: cool too slowly and cells die from osmotic dehydration; cool too fast and cells die from intracellular ice. A narrow "optimal" cooling-rate window minimizes both effects — but that window is only reachable for small, well-mixed samples like isolated cells or thin tissue slices, not a solid three-dimensional organ with a wildly heterogeneous internal cooling profile.
Simple cell suspensions — sperm, oocytes, blood, some embryos — have been cryopreserved successfully since the 1950s-1980s because every cell sits in near-direct contact with the surrounding medium, and heat can be added or removed almost instantaneously and uniformly across a microliter-scale sample.
A whole organ is a different problem entirely: it can weigh 150 g (kidney) to 1.5 kg (liver), with a core centimeters from any exchange surface. Cooling and warming propagate by conduction, and the rate achievable at the core is orders of magnitude slower than at the surface. Thermal gradients that are irrelevant in a droplet become catastrophic in a solid organ — the core can freeze into damaging ice while the surface is already glassy, and the resulting differential thermal expansion can crack the tissue outright.
This "cryopreservation size problem" is precisely why, until recently, only microliter- to milliliter-scale biological samples could be reliably banked — not the organs transplant medicine desperately needs.
The strategy that makes vitrification possible is deceptively simple: load the tissue with enough of the right chemicals that ice simply cannot form, no matter how cold it gets. In practice this means perfusing the organ's own vascular tree with a cocktail of cryoprotectant agents (CPAs) at concentrations far higher than any drug dose — high enough to be intrinsically toxic, which is the central engineering tension of the entire field.
CPAs fall into two broad classes. Permeating CPAs — DMSO, glycerol, propylene glycol, ethylene glycol, formamide — are small enough to cross cell membranes. Once inside, they hydrogen-bond with water, disrupting its ability to organize into an ice lattice, and colligatively depress both the freezing point and, more importantly, the nucleation temperature of the solution.
Non-permeating CPAs — sucrose, trehalose, dextran, hydroxyethyl starch, and synthetic "ice blockers" such as X-1000/Z-1000 polyampholyte polymers or antifreeze glycoproteins — stay in the extracellular space. They control cell volume during loading and unloading, and some directly adsorb onto nascent ice-nucleus surfaces, physically blocking crystal growth even at micromolar concentrations.
Delivered together as a multi-component cocktail, these agents work synergistically: each single agent can be kept below its individual toxic threshold while the mixture as a whole still reaches the very high total concentration — typically 40–60% w/v — required to vitrify.
Because CPAs must reach every cell in a solid organ, they travel through the organ's own natural highway: its vasculature. A cannula is placed in the primary artery (or vein) and CPA solution is perfused at controlled pressure and flow, following the same capillary network that once delivered blood.
Loading proceeds stepwise, ramping CPA concentration upward (often while cooling) over 30–90 minutes rather than in one sudden exposure. This graded approach exploits the fact that CPA toxicity is strongly time- and temperature-dependent — cells tolerate a brief exposure to near-lethal concentration far better than a prolonged one, and chemical toxicity itself slows sharply as temperature drops. Perfusing while cooling therefore "buys" tolerance for higher concentrations.
Uniform perfusion is never guaranteed: any region with reduced flow — a partially occluded vessel, an edge zone, fibrotic tissue — under-loads with CPA and stays vulnerable to ice formation during subsequent cooling, exactly there.
The two most cited historical cocktails — VS55 (~55% w/v: DMSO + formamide + propylene glycol) and the less-toxic DP6 (~6 M, egg-yolk-free) — were both engineered specifically to push whole-organ vitrification from theory into demonstrated rabbit kidney transplants.
This is the defining compromise of organ cryopreservation: the CPA concentration required to reliably suppress ice throughout an entire organ volume sits very close to the concentration that begins to damage cells chemically and osmotically. Push concentration higher and vitrification becomes more reliable, but viability drops from toxicity; push it lower and cells tolerate the exposure better, but the risk of at least partial ice formation during cooling or warming rises sharply.
Modern strategies attack this tradeoff from several directions at once: cocktails of several CPAs so no single one reaches its own toxic threshold, synthetic ice-blocking polymers that allow a lower total CPA concentration for the same anti-nucleation effect, loading at progressively colder temperatures where chemical toxicity is suppressed, and — critically — pairing high CPA loads with ultra-fast, uniform cooling and rewarming so less "safety margin" of concentration is needed in the first place.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| VS55 | DMSO + Formamide + Propylene glycol (~55% w/v) | High total CPA load, extensively used in early whole-organ vitrification studies | Proven vitrification of rabbit kidneys |
| DP6 | DMSO + Propylene glycol (~6 M, lower total load) | Reduced-toxicity cocktail, often paired with ice blockers | Better post-thaw viability margin |
| Ice blockers (X/Z-1000, AFGPs) | Synthetic polyampholyte or antifreeze (glyco)proteins | Adsorb onto ice-nucleus surfaces, block growth at µM concentration | Allows lower total CPA % for same protection |
| Non-permeating osmotic buffers | Trehalose, sucrose, HES, dextran | Control cell volume during loading/unloading, extracellular stabilization | Reduces osmotic-shock injury |
Vitrification is a physical trick that lets a liquid become a rigid solid without ever assembling a crystal lattice. As the CPA-loaded organ cools, rising viscosity progressively slows the molecular rearrangement needed to nucleate ice, until — below the glass transition temperature — molecular motion is effectively frozen mid-arrangement. The result is an amorphous solid: rigid, optically clear like glass, with no ice-water interface anywhere to damage a membrane.
Ice crystallization is thermodynamically favorable well below 0°C, but it is not instantaneous — molecules must find each other and self-assemble into an ordered lattice, a process with its own characteristic rate. Vitrification does not make ice formation thermodynamically impossible; it makes it kinetically impossible by starving the system of time and mobility.
As a concentrated CPA-water solution cools, its viscosity rises exponentially, following behavior close to the Vogel-Fulcher-Tammann relation. At the glass transition temperature — typically around -120°C to -139°C depending on the exact CPA cocktail and concentration — viscosity reaches roughly 10¹²–10¹³ Pa·s, high enough that molecular diffusion for crystal nucleation effectively stops on any biologically relevant timescale. Below Tg the material is a glass: amorphous, rigid, structurally frozen in time.
A glass at Tg is roughly a trillion times more viscous than room-temperature honey — molecules remain technically liquid-like in arrangement, but are so immobilized that the system behaves as a rigid solid for any realistic storage duration.
Vitrification succeeds only if the cooling rate is fast enough, relative to the CPA concentration present, to carry the tissue through the danger zone — roughly -5°C to -60°C, where residual free water and adequate molecular mobility make ice nucleation and growth most probable — before crystals have time to grow to a damaging size.
Higher CPA concentration raises viscosity at every temperature, independently slowing nucleation and buying tolerance for a slower cooling rate. Conversely, a sufficiently fast cooling rate can partly compensate for a lower, less-toxic CPA concentration, because there simply isn't time for ice to nucleate and grow even though it remains thermodynamically favorable. This is exactly why the "Cooling/Rewarming Rate" and "Cryoprotectant Concentration" controls in this simulation interact: a weaker CPA load can still vitrify successfully if cooling is fast enough, and vice versa — but push both too low simultaneously and crystallization becomes essentially certain.
Microliter droplets of CPA solution can be cooled at thousands of degrees per minute simply by plunging them into liquid nitrogen — fast enough to vitrify with relatively modest CPA concentrations. A whole organ cannot be cooled that fast throughout its volume: heat can only leave through the surface, and conduction through several centimeters of tissue limits the practically achievable cooling rate at the core to perhaps tens of degrees per minute — two to three orders of magnitude slower than a droplet.
This forces organ-scale protocols to rely on higher CPA loads, with their attendant toxicity, than would ever be needed for a cell suspension, and it means the coldest, slowest-cooling regions of the organ — typically the core — are the most vulnerable to residual ice formation: a spatial nonuniformity problem layered on top of the cooling-rate problem.
Once vitrified, the organ enters a state unlike anything in normal biology: a solid, amorphous, essentially motionless glass held far below any temperature at which chemical reactions or ice growth proceed at meaningful rates. This is the payoff of vitrification — a storage state in which time, for practical biological purposes, does not pass.
It is tempting to think simply keeping an organ cold — say, in an ultra-low freezer at -80°C — would preserve it indefinitely. It would not: -80°C is still far above the glass transition temperature of biological tissue (~-120°C or colder), so residual water retains enough mobility over months to years for slow ice recrystallization ("Ostwald ripening") to occur — small ice crystals gradually merging into larger, more damaging ones, driven purely by surface-energy minimization even without any temperature change.
True long-term stability requires storage below Tg, most commonly achieved in liquid-nitrogen vapor (roughly -150°C to -190°C) or, less often, in liquid nitrogen itself (-196°C). At these temperatures, molecular mobility is low enough that recrystallization, chemical degradation, and enzymatic activity all become negligible over any clinically relevant timescale — plausibly decades, with no established upper bound.
Below the glass transition temperature, essentially all diffusion-limited biological and chemical processes halt: enzymatic reactions cannot proceed because substrates cannot diffuse to the active site; membrane lipids cannot undergo phase transitions or peroxidative damage at any appreciable rate; free-radical reactions — a major driver of injury during storage of incompletely vitrified tissue — are essentially quenched.
This differs fundamentally from simple refrigeration, or even standard slow-freeze storage of blood and tissue banks, where slow degenerative processes continue at a reduced rate over months. A true vitrified glass is, for practical purposes, in biological stasis — the organ that goes into storage should be structurally and biochemically identical to the organ that eventually comes out, whenever that is.
Vitrified glasses are mechanically brittle, especially near and below Tg — large temperature gradients during cooling or handling, or simple thermal cycling and vibration during storage and transport, can generate enough internal stress to crack the organ, sometimes catastrophically. This "fracture problem" is now considered the major outstanding physical challenge for large-volume vitrification, distinct from the ice-nucleation problem addressed by CPA loading and cooling-rate control.
Current research directions include tailoring CPA cocktails and cooling profiles specifically to minimize thermal-stress accumulation, mechanically supporting tissue through the coldest transitions, and exploring alternative approaches such as isochoric (constant-volume) cooling chambers, which can suppress ice formation at less extreme, lower-stress temperatures.
Every gain made during cooling can be lost during rewarming: a vitrified glass that warms too slowly, or unevenly, will devitrify — nucleating and growing exactly the damaging ice crystals that vitrification was designed to avoid, now on the way back up rather than down. Solving rewarming at organ scale, uniformly and fast enough, has been the single largest barrier to clinical translation — and the recent breakthrough that has reopened the entire field.
As a vitrified glass warms, it must pass back through the same -60°C to -5°C danger zone it crossed during cooling — and on rewarming the risk is arguably worse. Ice nuclei too small to grow during the brief cooling transit can act as seed crystals during a slower warming transit, growing into damaging macroscopic ice if the tissue lingers in that range even briefly. This process, devitrification, can occur even in a sample that vitrified perfectly on the way down, if it is rewarmed too slowly.
Conventional rewarming — a warm water bath or warm air, relying on surface conduction inward — is fundamentally too slow for a solid organ of clinically relevant size, for the same conduction-limited reason surface cooling is too slow: heat must travel centimeters through insulating tissue, and the coldest, most vulnerable core is the last region to warm, and warms slowest of all.
Conduction-based warming of a solid organ produces a punishing tradeoff: fast enough to protect the surface almost guarantees the core lingers dangerously long in the nucleation zone, while slow enough to protect the core devitrifies the surface — there is no conduction-only rate that protects both simultaneously.
The modern solution abandons surface-in conduction entirely and heats the organ volumetrically, from within, using magnetic iron-oxide nanoparticles perfused throughout the vasculature — the same vascular network used for CPA loading — together with an external alternating, radiofrequency-range magnetic field. The nanoparticles absorb energy from the oscillating field and convert it to heat directly at their location, so heat is generated nearly simultaneously throughout the entire tissue volume rather than conducting in from the surface.
Because heating occurs near-simultaneously everywhere the nanoparticles reach — i.e., everywhere the vasculature reaches — warming rates of hundreds of degrees per minute become achievable even in solid organs of clinically relevant size: fast enough to cross the danger zone before devitrification nuclei can grow, and uniform enough to avoid the destructive thermal gradients that cause fracture.
In a landmark 2017 study (Manuchehrabadi et al., Science Translational Medicine), nanowarming rewarmed vitrified rat kidney tissue and vascular grafts at rates exceeding 100–200°C per minute, avoiding both ice damage and the cracking seen in conventionally rewarmed controls — a proof of concept widely credited with reviving serious research investment in whole-organ vitrification.
Even a perfectly rewarmed organ still contains the near-toxic CPA concentration it was loaded with, and this must be removed before the organ can be assessed or transplanted. Washout is loading run in reverse: the organ is perfused with a graded series of decreasing CPA concentrations, often paired with non-permeating osmotic buffers, to pull CPA back out of cells slowly enough to avoid osmotic shock — cells that re-absorb water too quickly as extracellular CPA drops can swell and lyse, undoing the protection just achieved.
Only after successful rewarming and washout can the organ actually be evaluated: histology, metabolic assays (ATP content, oxygen consumption), functional perfusion testing, and — the ultimate test — transplantation with recipient monitoring. Every step compounds: a failure at nucleation, cooling rate, rewarming uniformity, or washout osmotic control can each independently ruin an otherwise successful protocol.
Today, donor organs face a brutal cold-ischemia clock: viable preservation on ice — chilled, not frozen, with ongoing slow metabolic injury — is typically limited to roughly 4–6 hours for hearts and lungs, 8–12 hours for livers, and 24–36 hours for kidneys, even with modern machine-perfusion extending some windows further. This clock drives nearly every operational constraint in organ transplantation: frantic logistics of matching, transporting, and transplanting across a narrow window; thousands of viable organs discarded each year purely because a compatible recipient could not be reached in time; and the impossibility of building any real organ inventory.
A clinically validated whole-organ vitrification and nanowarming pipeline would convert transplantation from an emergency logistics sprint into a scheduled, elective procedure. Organs could be banked indefinitely, matched by full genetic and immunological compatibility rather than geography and time pressure, shipped by ordinary freight rather than emergency courier, and stockpiled against demand the way blood products already are — potentially eliminating a substantial share of waitlist deaths driven not by organ scarcity itself, but by the mismatch of timing and location.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Fahy et al., rabbit kidney | VS55 vitrification + conventional rewarming | First vitrified whole kidney to support life as sole functioning kidney after transplant | Proof that vitrified whole organs can retain function |
| Manuchehrabadi et al. (2017) | Nanowarmed rat kidney tissue / vascular grafts | Iron-oxide nanoparticles + RF field, >100°C/min uniform warming | First fracture- and ice-free rapid rewarming at scale |
| Follow-on rodent/porcine studies | Larger volumes, refined nanoparticle clearance | Scaling nanowarming toward clinically relevant organ sizes | Incremental progress toward large-animal feasibility |
| Clinical whole-organ vitrification | Not yet achieved (as of 2026) | Active research target across academic & biotech groups | Would enable indefinite organ banking if achieved |