"Realistic" Scenario
for Nanotechnological Repair
of the Frozen Human Brain

Reprinted from Cryonics: Reaching for Tommorow, Alcor Life Extension Foundation, 1991.

Abstract

Guidelines are suggested for designing realistic and defensible nanotechnological repair scenarios for the frozen human brain. A scenario is developed which is based on a) replacing brain ice with repair networks below Tg, b) carrying out gross structural repairs at temperatures in the range of about -100 to -30 degrees C, and c) carrying out most intracellular repairs at more elevated temperatures, relying in part on ordinary biological self-assembly and self-repair for carrying out much of the work required. The presently suggested scenario is intended as a rough outline that can facilitate rational discussion of the feasibility of repair. No mathematical analysis is attempted in this first specific description of "realistic" approaches to the repair of the frozen human brain.

Introduction

I. Definition

"Realistic" scenarios for repair are defined here as scenarios that might actually be applied, with appropriate modifications, to the restoration of the brains of patients in cryonic suspension. These may be distinguished from general proofs-of-principle (1,2,3) that attempt to demonstrate general feasibility without considering documented biological problems in detail, or that present the limits of the possible without considering what is most efficient, practical, and likely. No "realistic" repair scenarios have previously been proposed to the knowledge of the author.


II. Desirable Attributes of Repair Scenarios

Scenarios for repair of the frozen human brain should satisfy a number of important requirements. Although the scenario proposed here is based on the following guidelines, it does not include a self evaluation of feasibility (see "Testability" below) and does not attempt to be a fully-developed and fully documented work.

Factual Basis. First, realistic repair scenarios must be based on what is known or can be inferred about the nature of the actual injury present in frozen brains and frozen brain tissue. This is mandatory because, by the definition above, a realistic scenario attempts to set forth specific approaches to solving the problems of repair, which is impossible if the problems are not clearly and accurately identified. It is essential to avoid addressing problems that do not exist and to avoid overlooking important difficulties that are likely to arise. Focusing on real problems is perhaps the best way to realize a technically compelling result.

Parsimony. Second, the repair scenario should be parsimonious. It should not attempt to do more than the minimum amount of work required for a satisfactory result. The reason for this is that smaller jobs are more easily and more credibly solved than larger ones, and the goal of a repair scenario is to demonstrate the feasibility of repair. The fewer the number of tasks required for repair, the more likely it is that these tasks can be accomplished. This of course does not mean that any real problem should be minimized or ignored, but it does mean that needless tasks (such as keeping track of individual sodium atoms) should not be considered. Parsimony requires distinctions to be made between what is important and what is not important.

Detail. Third, parsimony in selecting problems to solve does not necessarily imply parsimony in describing the details of the problem solving process. Too much detail is likely to be confused with fortune telling, which would be inappropriate. However, when the appearance of fortune telling can be avoided, hard detail lends reality to the scenario, particularly if it is backed up by numbers. Thus, rather than just saying, for example, "the fractures will [somehow!] be removed from the data base," it would be more plausible to spell out in detail a defensible molecular plan of attack on the problem.

Testability. Fourth, of course, the repair scenario must be physically achievable. Ideally, it should be possible to evaluate each step of the repair scenario quantitatively to test its feasibility. Each step should also be specific enough for a more general evaluation of its feasibility to be made. In short, the scenario should be testable and falsifiable in all ways possible. A recent criticism by Fahy of previous discussions of repair of frozen brains (4) focused in part on such evaluations, e.g., evaluating the feasibility of providing power for nanoprocedures at cryogenic temperatures. Although the latter criticism and many similar points will not be addressed here, a consideration of such details is important for fully-developed scenarios.

Defining The Problem: Freezing Damage

I. What Is and Is Not Considered

The repair scenario described here deliberately ignores extraneous problems such as postmortem damage, transport injury, circulatory obstruction, and previous traumatic brain damage or other devastating types of cerebral deterioration because these problems represent side issues and are not inevitable given the level of hospital cooperation that has frequently been achieved. Furthermore, legal changes could allow freezing to be carried out under much better conditions, so that fewer such problems will arise. Other types of repair scenario should if necessary (and possible) be devised to address these problems separately. However, the damage caused by cryoprotectant perfusion prior to freezing is discussed.

The repair scenario will assume conventional preservation, i.e., the brain is not fixed prior to freezing: repair scenarios for fixed brains are likely to be substantially different from what follows, for substantially different problems will be present. It will also be assumed that the "Smith Criterion" (5) is attained or bettered: grossly inadequate concentrations of cryoprotectant will produce such massive mechanical damage (6) that repair, if it is possible at all, may well depend on different principles than those described here.

The discussion that follows in the next two sections is necessarily based on limited information, some of which may be misleading. It goes without saying that more research is needed on every point discussed. The incompleteness of present information, however, does not appear to be sufficient to preclude meaningful evaluations and the development of reasonable repair scenarios. The format of the next two sections will consist of a series of statements describing potential problems followed in each case by an evaluation of the problem's likely relative seriousness.

II. Perfusion Damage and its Significance

Rabbit brains fixed after perfusion with glycerol at temperatures above 10-15 degrees C and examined histologically do not show loss of ground substance or substantial morphological alterations (6). Glycerol perfusion under low temperature conditions can, when concentrations are high (6 M), cause extensive shrinkage of the brain as a whole and of the component cells and processes, with distorted histological staining (7, unpublished observations).

Evaluation. Cell shrinkage may cause problems similar to those that occur during freezing, and will thus be considered in more detail below. Perfusion, unlike freezing, however, may remove proteins and cellular debris from the brain. Unless enormous concentrations of glycerol are used, however, extensive protein loss from previously undamaged brains appears unlikely. Altered staining implies, at worst, altered chemistry, particularly since staining is done when, presumably, all glycerol has been removed from the tissue. However, structural preservation of such brains is apparent (6,7), and it is quite possible that altered staining is the result of unusual fixation in the presence of glycerol or similar artifacts. If chemical changes have taken place as a result of glycerol exposure, these changes should be reversible due to their stereotypical nature and the identifiability of the chemically modified sites: actual information loss is not likely. Moreover, these chemical changes do not appear likely in the usual case, in which brains are perfused with lower concentrations of glycerol than 6 M.

Some chemical alterations by glycerol are likely in even the best situations. These alterations, however, primarily should be altered levels of ordinary metabolic intermediates due to the actions of enzymes on glycerol (8) or due to the differential effects of glycerol on the kinetics of different enzymes (9). As such, they are fundamentally trivial and close to being spontaneously reversible. Glycerol has apparently never been documented to denature any protein under any conditions, with the possible exception of glycerol in concentrations in excess of 95% w/w (10), a condition not remotely approached in cryonic suspension, even during the freezing phase of the process (11).

III. Freezing Damage: Significance of Different Types of Damage

1. "Biochemical/biophysical" Freezing injury. Imagine the appearance of a "frozen" brain cell. Approximately 60% of the volume of the brain has been converted into extracellular ice (5). Freezing has extracted large fractions of the intracellular water and thereby reduced cell volume (12). This in turn has reduced cell surface area, which has the potential of forcing an expulsion of membrane material from the plane of the membrane (see below). The combination of cellular shrinkage, lowered temperature, and elevated glycerol concentrations may cause the following kinds of damage.

a) Extrusion of pure lipid species from the plasma membrane, either on tethers (13,14) or as free lipid droplets in the cytoplasm (15) or in the extracellular space (proportional to the reduction in membrane surface area produced by freezing).

Evaluation. This is a phenomenon seen so far only in plants. Shedding of lipid to free-floating extracellular droplets has not been seen even in plants. As long as any lipid extrusions are intracellular or still attached to the cell of origin, it should be clear where to redistribute the lipid on warming, if necessary. The main difficulty arises from the inability of lipid extruded in this way to spontaneously return to the plane of the membrane during volume expansion on thawing: restoration of approximately isotonic volume near the melting point causes cellular lysis in plant cells (13-15) due to inadequate membrane surface area. This should be a relatively easy problem to address and does not involve appreciable information loss on freezing.

b) Loss of membrane proteins (possibly including hormone receptors and potassium channels) into the extracellular space.

Evaluation. Loss of glutamate receptors has been documented in brain tissue frozen without cryoprotectant (16). The number, state, and precise anatomical distribution of potassium channels in hippocampal dendrite membranes probably encodes memory to a large degree (17), so the potential loss of these membrane proteins is of concern. However, significant neurotransmitter receptor loss has only been seen when no cryoprotectant was used at all, in two papers in the literature (16). In all other cases (18), even when only low concentrations of extracellular cryoprotectant (sucrose) were present (19), all functions tested have been present, implying proper retention of membrane proteins. Even intracellular organelles do not redistribute/mix proteins to a worrisome degree after brain tissue is frozen and thawed (20). Thus, even excessive cellular shrinkage prior to freezing (caused by inadequate penetration of glycerol) superimposed on subsequent freeze-induced cell shrinkage should not subject cells to greater osmotic stress than has been shown experimentally (by freezing with only extracellular sucrose present as cryoprotectant [191) not to cause major loss of membrane proteins.

Beyond this, however, are several other supportive observations. First, associative learning involves not just potassium channels but also changes in several other characteristic proteins that induce and maintain the alterations of potassium conductance underlying memory (17). Even loss of potassium channels should leave these remaining proteins behind, providing a clear indication of the "trained" vs "untrained" state of given dendritic synapses and/or perisynaptic regions. Second, altered K+ permeability probably involves durable chemical modification of the potassium channel, so that lost channels could be identified as "trained" or "untrained" and counted as such; improper return of individual molecules to specific synapses would likely be irrelevant as long as the proper total number of "trained" potassium channels ends up at each "trained" synapse or perisynaptic region. Finally, durability and inferrability is further implied by the associative nature of memory, in which a given memory is stored redundantly in several brain regions in a number of independent forms (17,21), all of which are unlikely to be extinguished simultaneously in their entirety.

c) Denaturation of proteins.

Evaluation. Likely to apply, at most, to very few proteins (22). Furthermore, protein denaturation is inherently reversible (23). An apparently trivial issue.

d) Improper disulfide bridge formation between some proteins (24).

Evaluation. Also an almost negligible problem, for similar reasons.

e) Leakage of concentrated extracellular solute into brain cells.

Evaluation. Like membrane lipid loss, the main problem caused by this leakage, should it occur, would be expansion-induced lysis on thawing (25). Although the volume of extracellular space in the brain might be considered insufficient to permit lysis on warming, ultrastructural evidence of disruption of neuronal fine processes in frozen-thawed brain (26) as well as ultrastructural evidence of swelling of frozen-thawed synaptosomes (27) lends credibility to this possibility. However, this problem can be handled in principle even more easily than the lipid loss problem, simply by extruding the extra intracellular osmolyte during thawing. Not a significant problem.

f) Local leakage of brain cell solute to the extracellular space. This problem can be subdivided into leakage of small ions (primarily potassium) and small metabolites on the one hand and large metabolites and proteins on the other.

Evaluation. The former problem should be negligible; pumping potassium back into neurons should be straightforward (given non-leaky membranes: freeze-permeabilized membranes evidently reseal during warming and thawing [281) and small metabolites can be resynthesized from supplied nutrients. Leakage of proteins and other large molecules is more serious. However, significant (e.g., 50%) uncorrected loss of intracellular soluble protein from cell bodies could probably be sustained without creating very serious problems, since it should be possible to institute compensatory controls over metabolic rate and membrane permeability consistent with the spontaneous ability of the cell to resynthesize missing proteins on warming. Even massive protein loss from cell bodies would not be able to erase cell identification, since cell identification will be encoded in the types of synapses the cell makes (which will be preserved [181), by the pattern of genetic expression readable in the nucleus (which will probably also be preserved [291), and by membrane and perhaps non-soluble cytoplasmic protein markers (which will be preserved well enough [201).

A different kind of potential problem could result from protein loss from torn axon bundles (6,30). Severe losses of axoplasmic proteins at sites of tearing could make the identification of individual nerve fibers on both sides of tears more difficult, which may limit the ability to deduce the original connectivity of the brain. However, it is likely that a short distance away from the tear the axonal protein content should be largely unaffected by the tear, especially given the gel-like nature of axoplasm (31) and the relatively rigid structures mediating axoplasmic traffic.

Significant loss of proteins would undeniably complicate the restoration of metabolism significantly, and should be avoided to the extent possible: returning proteins to their proper sites would be comparatively difficult, and simply correcting for the losses as referred to above requires considerable metabolic "tinkering." Loss of non-proteinaceous larger metabolites once again could significantly complicate neuronal identification and would require considerable "Humpty Dumpty" work that would be best to avoid if possible. During freezing, such damage will be somewhat limited by the extracellular diffusion barriers presented by ice and high viscosity extracellular media, but during thawing the extracellular space will be progressively "stirred" by declining viscosity, thermal expansion, convection, and cellular expansion. Means of blocking this "stirring" would thus be important to deploy.

g) Precipitation of proteins and cellular buffers.

Evaluation. Little (32) direct evidence exists for this mode of injury; if it were to occur, the result would be reduced metabolic competence secondary to denatured or missing proteins or unfavorable pH's for normal metabolism. The remedies--supplying replacement buffer and/or proteins and restoring precipitated proteins and/or buffers to a soluble condition--seem fairly easy to deal with.

h) Leakage of lysosomal enzymes into the cytoplasm, predisposing to intracellular autolysis on warming (33).

Evaluation. Not demonstrated to occur. Glycerol and low temperatures can be expected to limit autolysis during cooling, and exogenous inhibitors should be able to control autolysis during warming. Not a serious problem.

i) Reorganization of membrane bilayer structure into HexII forms, i.e., cylindrical lipid tubes (34,35). This change is spontaneously reversible, in part, upon warming and rehydration, but will keep the membrane leaky in the cold. Phase separation of lipid subclasses within the membrane, producing leaks secondary to the resulting molecular packing faults in the membrane (36,37).

Evaluation. These are problems functionally comparable to osmotic or mechanically-induced leaks noted above. No direct evidence for HexII transitions exists for any mammalian system, and HexII forms appear unlikely in the presence of 3-4 M glycerol before freezing, since HexII is a dehydration form (35), and glycerol can prevent the required level of dehydration for HexII formation from taking place (11). In both HexII formation and more conventional phase separations, all membrane material remains in the membrane. The problem then becomes one of preventing additional leakage from taking place during thawing, and of redistributing solutes across membranes as needed after membrane resealing is completed. This seems achievable in principle. In the case of HexII, spontaneous reversal of the phase transition may lead to incorporation of lipid and protein into incorrect leaflets of the membrane. Controlled reversal, however, should be able to direct proper redistribution rather easily.

j) Breakdown of the structure of cytoplasm into blobs of proteinaceous material (38,39).

Evaluation. This may occur if there is a breech of the cell membrane or for other reasons that are so far poorly understood. So far, this phenomenon has been seen only in kidney, not in brain, and does not pertain to all cells, even in the kidney. If it should occur in brain, repair could be complicated, but it is doubtful that any actual information loss would occur. This change might be spontaneously reversible on warming.

2. Mechanical freezing injury. The most pressing kinds of damage are mechanical forms of damage, not only because it is this type of injury that has actually been observed in frozen- thawed brains, but also because the potential for actual information loss is much more serious than is the case for the biochemical challenges just considered. It is not yet certain that high concentrations of glycerol will prevent such injury consistently.

Several kinds of mechanical injury could occur, including the following.

a) Memory may be encoded in part in the shapes of dendritic trees (17); these shapes might be altered by freezing.

Evaluation. Dendritic remodeling associated with learning seems to involve massive changes such as deletion of unused synapses and unused dendritic branches (17). This remodeling is associated with the actions of many different proteins and, most likely, with considerable changes in remaining synapses (17). It is likely to be not the shape of the dendritic branchings that is important but, instead, the specific pattern of connections, and this pattern in turn is presumably responsible for the changes of shape of the dendritic trees (17). Thus, the shape changes induced by freezing and thawing should be irrelevant as long as the synapses and dendrites remain physically intact. Freezing is well known to spare synapses (18).

Although there is evidence for axonal (6) and possibly cellular (30) tearing, light microscopic evidence suggests that well-glycerolized hippocampal dendrites are not broken by freezing and thawing (6). But even considerable freeze-induced damage to dendrite branches might still leave the pattern of connections obvious from the remaining synapses. Dendrite branching patterns and their underlying biochemical correlates are biologically robust and dramatic and should retain a high degree of inferrability, particularly if "stirring" is avoided during warming.

b) Disruption of non-synaptic junctions between cells and capillary separation from the surrounding brain tissue.

Evaluation. Such problems have been observed (40), but would appear not to involve direct information loss. Such separations could tear fine processes, however, so the imperative to prevent extracellular diffusion on warming is reinforced again by such observations.

c) Local (not global) ripping, twisting, and fraying of the ripped ends of nerve tracts by contraction of the brain cells and by the push of extracellular ice, creating debris-strewn gaps measured in microns in both length and thickness (6,30,40).

Evaluation. A severe form of damage. Reconstruction may require a certain degree of luck, i.e., the existence of positional relationships between nerve fibers in a given tract that do not vary significantly from one side of the gap to the other. Should such consistent positional relationships exist, inferring the proper connections at the site of a gap should be straightforward. However, if the positional relationships happen to be changing at the point of such a gap, additional information in the form of molecular markers that might identify individual fibers may be required for accurate inference of the pre-existing connectivity. Electrical tests across the gap may also be required to check for physiological consistency.

Should molecular markers be identical from one fiber to another, should positional relationships prove unreliable, and should electrical tests prove ambiguous, enough information for correct reconstruction may be present in the debris pattern that exists in the frozen state, considering the limited opportunities that exist for diffusion during freezing. In this event, prevention of "stirring" of the debris during warming will be critical. Finally, reconstruction might be possible based on consistent, minute size differences between fibers. Should all of these sources of information fail, however, the infrequency of these gaps and the generic effects of many connections as well as the vast redundancy of the brain may make incorrect inference of the proper connections still consistent with an adequate ultimate clinical outcome. (Clinical observations suggest that severe local damage can be consistent with maintenance of identity and personality.)

d) Fracture and separation of fractured halves of cells, axons, dendrites, capillaries, and other brain elements by distances in the millimeter range after the temperature drops to below the glass transition temperature (40,41). (Observations of the gaps referred to in c) above might also reflect microfractures that became "mushy" on warming and thereby resulted in molecular blurring of the fracture faces.)

Evaluation. A catastrophic form of injury, offering perhaps the greatest challenges for the design of molecular repair devices. However, this injury in itself may involve little or no actual loss of information. This is the most non-physiological type of injury and will require the most radically innovative types of repair system for its reversal. Such repair systems do, however, seem inherently possible.

e) Physical disruption of capillaries due to intracapillary ice formation: rupture of capillary wall, tearing of endothelial cells, stripping of endothelial cells from their underlying capillary wall material, resulting in incompetent vessels littered with emboli (40).

Evaluation. A very serious form of injury. However, no information is contained in capillaries per se. The entire capillary network could likely be cleared out and replaced with generic capillary "transplants" without any effect on the identity of the patient. Repair of the existing capillaries would require innovation on the order of what would
be required to repair fractures. Reparable in principle.

f) Stripping of myelin from axons (40): formation of gaps between the axon membrane and the myelin, unravelling of the myelin, possible tearing of the axolemma resulting in loss of intra-axonal material at moderately low temperatures.

Evaluation. Myelin is inert, generic, non-information-containing material. Despite the types of myelin damage described, there should be no problem in inferring which regions of axolemma were previously covered by myelin and which were exposed. Myelin's function is only important under physiological conditions. Myelin repair might therefore not be necessary until the patient was restored to normal body temperature, at which point it could probably be carried out by ordinary or modified oligodendroglial cells, which lay down myelin under normal conditions. Leakage of axonal material has been considered above; it may be reduced by the presence of even a tattered myelin sheath which would act as a diffusion barrier.

Defining The Problem: Constraints on Repair

Repair scenarios must recognize that some kinds of repair would be extraordinarily difficult, futile, or even counterproductive to carry out at the lowest, most protective temperatures. For example:

I. Osmotically-induced Cellular Shrinkage

Extruded lipids and proteins cannot be reinserted into the membrane until the cell volume is once again increased because there is no room for them. Restoring cell volume while the cell is in the vitreous state would be many orders of magnitude more difficult than performing the same process at higher temperatures, and would be a seemingly ridiculous and possibly even impossible task to attempt.

II. Phase Transitions

Low temperatures and membrane dehydration per se cause membrane lipid species to crystallize or undergo HexII reorganizations. This is therefore the natural state of these lipids at the prevailing temperatures. Any attempt to reorder the membrane lipids into a lamellar phase will lead to spontaneous re-separation of these phases either at the prevailing temperatures or on warming. Thus, simply "repairing" this membrane defect at cryogenic temperatures would be futile. Introduction of alien lipid species to prevent re-separation would be problematic due to the absence of room in the membrane for such species and the need to subtract native lipid to make room. These changes would all have to be reversed later, and might create more problems than the original phase separations.

III. Denaturation

Any denatured proteins will also prefer to be denatured under the prevailing conditions. Renaturing them will only lead to re-denaturation as temperatures inevitably rise later on. Preventing re-denaturation would require special "chaperones" for each protein, whereas waiting for most denatured proteins to spontaneously renature (23), in part or completely, during warming would avoid most of the need for such artificial molecular folding-control devices.


IV. Changes in Tissue Volume: Thermal Expansion, Brittleness, & Elasticity

A fracture represents anisotropic contraction of cerebral tissue due to temperature reduction or inhomogeneous expansion during warming. Local rips in axons may arise for similar reasons. To fill in gaps caused by the inherent thermal contraction of cerebral tissue may create a problem when the temperature is raised and all of the existing structure, both the native structure and the added structure, is inevitably forced to expand: expansion lesions such as buckling and shearing of axons may replace the previous contraction lesions. It may be wiser to allow thermal re-expansion during warming to at least partially close these gaps and to effect repair only after this happens.

Likewise, many axons may be very stretched. Destretching them by adding material to them could cause the same buckling problem when warming occurs. Finally, tissue will be brittle below TB and may be brittle even at temperatures moderately above this. Physically moving structures around under such conditions may damage them. Thus, attempting to close a fracture by physically forcing the two sides together is liable to rip structures on both sides of the gap. Thus, some repairs made below To could induce the need for more repairs later when the temperature is elevated.


V. Changes in Tissue Functionality

Statements have been made in the past to the effect that various cell structures, e.g., mitochondria, will be restored to a "functional state" while still frozen (3). This would, however, represent a nonsensical goal for many reasons, not the least of which is that functionality requires dilute aqueous liquid solutions, which cannot exist at low temperatures. The correct goal is to ensure that function resumes after warming to physiological temperatures, regardless of the repair pathway that must be followed during warming from lower temperatures to attain this goal.

The Repair Scenario

We will assume that the repair procedures begin at a temperature slightly below the glass transition temperature of the system.


I. Stabilizing Fractures

The first step is to stabilize existing fractures. Fractures require special treatment, and they require it from the very beginning since, as we will see shortly, the second repair step will obliterate non-organic components of fracture faces and will thus make it more difficult to match fracture faces and guide these faces together later if special precautions are not taken at the outset.

So, the very first step is to infiltrate surface fractures with specialized molecular devices which will form coatings or surface replicas of the fracture faces to molecular or near-molecular resolution. (Note that the process of fracturing releases energy that creates a very high though very brief local elevation of temperature. The first several molecular layers on each side of a fracture may therefore be somewhat melted or disordered. Therefore, absolute molecular resolution may not be attainable.) It is known from standard freeze-fracture microscopy that fracture faces can be coated below Tg with metal films that retain their structural fidelity even after the tissue is dissolved in Chlorox!

Thus, the formation of sufficiently stable fracture face replicas at temperatures below Tg appears feasible and would maintain the overall geometry of the fracture faces after dissolution of the portion of the face that is ice and glass. Pores in the replicas of areas of pure ice or pure glass should be included to permit outgassing during the subsequent sublimation process (see below), which otherwise could tear holes in the replicas.

After coating of opposite fracture faces, these faces could be computationally compared to verify complementarity. After complementarity analysis, the repair system could build filaments between the faces. The filaments on each side of the fracture would be complementary to each other and would connect so as to maintain fracture face registry later when the temperature is raised. Given sufficiently strong replicas, these "guide wires" could be attached only to the replicas (the replicas in turn being tightly adherent to the fracture faces themselves at all points).

The function of the wires later would be to direct each fracture face as a whole toward the other fracture face as the gap is later closed by normal thermal expansion in such a way as to continue to ensure perfect registry of the two fracture faces as the gap narrows. Molecular "ratchets" along the guide wires could apply small forces to encourage closing where this is necessary. If the "guide wires" are built onto the replica faces at the sites of special pores, then as the gap is closed and the faces approach each other, the "guide wires" can be allowed to protrude into safe regions of tissue on each side of the gap, and/or they could be disassembled at a pace set by the narrowing of the gap.

For deeper fractures not accessible from the surface, the same process might be accomplished by excavating the vascular compartment first, pausing for fracture stabilization as fractures are encountered.

II. The Need for an Overall Orientation

The next thing to do is to get the big picture. The frozen brain contains highly shrunken cells and neuronal processes compressed between sheets of ice and pools of vitreous cryoprotectant water-solute inclusions. There may be lipid extrusions, floating debris, ripped axons, hemorrhaged capillaries, stabilized fractures, unraveled myelin, crystallized regions of certain surface membranes, extruded cell contents in the extra-cellular space, and other relatively gross alterations. We desire to identify and stabilize all of these lesions before significant "stirring" is permitted. This is difficult to do without large-scale cooperation of repair devices, for which a coordinate system needs to be set up, preferably one that does not in itself cause any damage.

III. Excavating the Extracellular Space

We approach the problem by capitalizing on the fact that about 80% of the brain is nothing but water and cryoprotectant (42) and that most of this exists in the form of pure ice located in the extracellular space. We first desire to remove the ice and most of the vitrified extracellular solution. This step has two important advantages. First, it creates room for the deployment of an extracellular communications complex which will be used to direct subsequent repairs. Second, it makes transmembrane diffusion in either direction ("stirring") effectively impossible when the temperature is subsequently raised.

1. General Description of Method --Our task might best be accomplished by a combination of direct excavation (done by relatively stupid molecular "jackhammers"), which creates a certain amount of local warming, and by spontaneous ice sublimation, which offsets some of the local heating due to evaporative cooling. The rate of excavation is set so as to generate net local temperatures of around -120 degrees C (i.e., about 5 to 15 degrees C below the limiting To for glycerol water-solute systems).

Excavation might proceed by digging out hollow tubular "mines" through the ice perhaps 1 micron or more in diameter. The insides of the "ice mines" are maintained at a strong vacuum. In a vacuum simulating that of deep space, ice evaporation rates have been shown to be sufficient near Tg to move sublimation boundaries at rates on the order of microns per day (43)! If we maximize the surface area available for evaporation while also maximizing the rate of direct excavation, it might be possible to remove extracellular ice fairly rapidly-for example, in a few months.

In doing the excavation, we are not limited to entering through the vascular system. We can enter through ice channels wherever they may be, and they will be everywhere, and larger in extent than most biological structures. We can also enter through special ports built into fracture face replicas. The cerebrospinal fluid cavities can be evacuated with bulk technologies or a combination of bulk and molecular technologies. Since the entire brain is under a strong vacuum during this process, pressure gradients that could cause mechanical failures should be minimal.

In addition to the advantages of minimizing heating while maximizing water mobilization, sublimation is also advantageous in that it is self-limiting in a favorable way. Sublimation of water, as opposed to ice, raises the glass transition temperature in the sublimed region and thus halts further sublimation. Therefore, we can remove the ice by sublimation without excessively dehydrating the glassy matrix surrounding the biomolecules of the brain, either intra- or extracellularly. Nothing but water can evaporate off at these temperatures.

Nowhere in this scheme is it necessary to pay the slightest bit of attention to documenting the locations of water molecules or extracellular solutes such as glycerol, sodium, or chloride, or worrying about their orientations. Other extracellular material, such as debris, however, poses some problems. We will return to these problems momentarily.

2. More Specific Description of Molecular Excavators -- Two or more types of molecular "jackhammer" are envisioned. The first type is envisioned to attack only ice. The action of this excavator is to dislodge individual water molecules from ice and pass them to a "molecular conveyor-belt" system analogous to the conveyors used by axons to drive axoplasmic flow. The conveyor system transports water molecules to sites external to the brain. The second type of excavator removes vitreous material, such as glycerol, glycerol+water, and glycerol+salt+ water. These clumps of molecules are then passed to the molecular conveyor for transport outside of the brain.

Small molecules operating at temperatures near -120 degrees C cannot be self-powered. Therefore, these molecular devices must be attached to a power distribution source. One practical means of achieving this may be to attach the sensor-effector elements to a long mechanical rod which delivers the impulse required to disrupt the appropriate non-covalent bond once the sensing element identifies a proper target. This rod-tip association might be envisioned as a sort of "molecular steamshovel" in construction, with the ability not only to relay an impulse provided from the central power source, but also to position the effector tip in a versatile fashion using accessory positioning elements. The excavation could proceed with minimal, entirely local "computations" by following a stereotyped sequence of steps little more (and possibly less) complicated than the "computations" carried out by a ribosome.

In order for sublimation to proceed at the highest possible rate, collisions between sublimed ice and the ice vacuum interface should be minimized, since sublimed water will with some probability stick to that interface and require re-sublimation or direct dislodgement by the molecular evacuators. Without attempting to design an efficient means of proceeding with excavation so as to minimize this problem, it can be noted that simply designing the outgoing (but not the incoming) portion of the conveyor to be able to adsorb free water molecules from the vacuum would be helpful.

The energetics of both sublimation and molecular excavation should be reasonably easy to calculate. These processes might well be the most energy-intensive part of the repair process.

3. First Complication of Excavation: Avoiding Membrane Fracture-There is a fundamental problem of removing the vitreous residue from the extracellular space at temperatures just below To, namely, that the strengths of the non-covalent bonds holding together membrane bilayers are very much lower than the strengths of the non-covalent bonds holding together the vitreous matrix. In ordinary freeze-fracture microscopy, fracture planes often proceed along the plane of the middle of the membrane bilayer for this very reason (44). Applied to the problem of excavating the vitreous residue, this means that some method must be used to ensure that energy delivered to dislodge segments of the vitreous residue does not accidentally dislodge lipids from cell membranes.

A variety of methods might be brought to bear on this problem. One method might be to avoid regions that may contain nearby membranes as indicated by detection of membrane markers protruding a considerable distance into the extracellular maxtrix beyond the lipid bilayer proper. While this would lead to incomplete excavation, the opportunities for transmembrane diffusion might nevertheless be reduced sufficiently for this to be a satisfactory paradigm.

Another method might be to check obviously large dislodged chunks of residue for the presence of lipid and, if lipid is found, to separate it from most of the vitreous residue in which it is embedded and to reinsert it to its original site before proceeding. It might also be possible to design the geometry of the force application process so as to ensure that only a few molecules are dislodged at any one time, the force being applied not to the medium at large but to a very local and superficial area (e.g., the third vertex of a triangle).

4. Second Complication of Excavation: Extracellular Debris--There could be a considerable amount of extracellular debris. It is essential not to remove or damage this material, since it may be critical for inferring the undamaged state. Fortunately, the power applied per piston cycle need only be sufficient to break noncovalent, but not covalent, bonds, so extracellular debris should not be degraded by the excavation process. However, it will be necessary to fix all such debris in place, a nontrivial procedure.

Perhaps the best approach to this problem would be to erect side branches on the molecular conveyor belts. These side branches, shaped something like trees with their trunks originating on the conveyor belts, would possess binding sites and/or molecular clamps selected for the encountered debris and would bind all such debris noncovalently in place. The binding would be such as to represent the original three-dimensional distribution of the bound debris. For this type of molecular "book-keeping," it may not be necessary to completely strip the debris of surrounding glassy phase. In any case, the tree-like structure of the debris binding elements should ensure that all debris can be more-or-less locked in place during the repair process, thus permitting extracellular excavation to proceed without loss of information.

5. Third Complication of Excavation: Inadvertant Excavation--A further complication arising from this process is the possible "accidental" excavation of exposed cytoplasm/axoplasm. This can perhaps be avoided by solving the problem of excavating extracellular debris. Sensors for biological materials that permit immobilization of debris could similarly seal off cut axons and ripped cells.

As excavation concludes, the vacuum level should be reduced to ensure that additional unwanted sublimation of water (mummification) does not take place as temperature is later raised. The empty spaces can be filled with inert gas and/or with water vapor in equilibrium with the tissue at ambient temperature.


IV. Establishment of Extracellular Repair Network

As excavation/evacuation proceeds, an extracellular communication, transportation, and coordinate system could be laid down in the space made available. This system, penetrating throughout the extracellular portion of the brain and in intimate physical contact with the brain everywhere, could be thought of as a kind of "meta-brain," capable of relaying information about the brain over long distances while potentially having a volume amounting to more than 60% of the original volume of the entire brain (which is roughly the volume previously occupied by ice). This volume represents about 150% of the volume of the cellular components of the brain. The metabrain would permit all exposed lesions to be mapped and analyzed. Undamaged structures could be passed over without further action, except as they are needed to infer the proper locations of abberant structures, such as debris resulting from ripped axons. Furthermore, the metabrain could be in contact with external computers, where most computation might occur.

V. Repair Computations for the Extracellular Space

At this point, all labile extracellular structures have been physically immobilized and a coordinate system is in place. No "stirring" has taken place because all procedures have been carried out just below Tg. All significant extracellular anatomical elements of the brain have been registered. The "wiring diagram" of the brain can now be deduced, and all damaged areas can be catalogued as to type and place. Where necessary, the loci of missing structures could be deduced at this point. For example, ripped bundles of axons are analyzed to deduce how to infer the pattern of connections between the two ripped ends based on the direct physical remains of the ripped axons and any other available information. The loci of missing cell membranes are deduced. Extracellular debris are assigned to appropriate destinations.

To this point, no actual repairs have been made and the process has been completely noninvasive as far as the actual cells of the brain are concerned. Based on the results obtained to this point, specialized repair devices are assigned to specific tasks and specific regions.

No tasks so far have involved the making or breaking of covalent chemical bonds. All excavations, sensing, and computations have been based on purely physical processes which should be able to operate at cryogenic temperatures given an adequate external power source and power transmission system.

VI. Warming above Tg

In order to proceed with repairs, warming of the brain is slowly induced. The advantages of warming are several. It induces changes in volume which permit fracture healing, it induces desirable changes in tissue pliability/deformability needed for moving structures such as cell membranes, and it permits both diffusional transport of needed molecules and the chemical reactions needed for repair.

The primary hazard of warming is not biochemical but diffusional. At temperatures as high as about -50 degrees C, virtually no enzymatic activity should be possible (22, 45). Deterioration at this temperature is likely to be due to slow intracellular diffusional processes perhaps accompanied by slow spontaneous breakdown of certain relatively rare labile molecules. Any enzymatic activity that could occur is likely to be arrested in time due to lack of available substrate or accumulated product inhibition, and thus is unlikely to proceed very long. A special class of proteins, catabolic enzymes, may pose the most serious problems. However, the fraction of enzymes represented by key catabolic enzymes is small and all such activity can be blocked by specific inhibitors.

We can at this stage also ignore problems arising from any protein denaturation that may exist. Denatured proteins are not likely to catalyze troublesome reactions and are not needed for any functional role, so there is no reason to worry about them until temperatures are brought to near-zero. At that point, many or most of them will have spontaneously renatured, or will renature spontaneously given additional warming. The remainder can be renatured and disaggregated specifically at temperatures near 0 degrees C using specialized devices for each labile enzyme. This process should be trivial enough to ignore for the present purposes, particularly since the number of denatured proteins in glycerolized frozen-thawed brains should be minimal as a fraction of the total.

Thus, the primary initial portion of actual repair, as opposed to simple survey and marking of the damage, consists of coping with diffusional processes. At temperatures between about -110 and -50 degrees C, two major types of diffusional process can be identified: the diffusional motions that blur the fracture interfaces we have previously marked and prepared for healing, and diffusional motions within cells. By removing the great majority of the extracellular space and immobilizing extracellular debris, we have precluded transmembrane and extracellular diffusion, and by forming durable fracture replicas and establishing the relationships between them, we have precluded diffusional information loss at these sites. Intracellular diffusion is relatively trivial over the short run. We therefore are able to proceed with the extracellular repair process first, and then to turn our attention to cellular interiors.

VII. Fracture Healing

Coefficients of thermal expansion and water/glycerol diffusion coefficients dictate the kinetics of spontaneous fracture healing in pure solutions. Extrapolation of available data for glycerol-water solutions suggests that spontaneous fracture healing in these solutions will first become appreciable in the vicinity of about -80 degrees C (46). Thus, we will want to heal fractures in cerebral tissue during warming from -100 degrees C to about -80 degrees C. The key issues involved at this point are a) the removal of the protective replica surfaces and b) the union of tissue on either side of the fractures. Both a) and b) pose significant problems. Removing the replica surface will tend to free bound species on each side of the gap for undesired diffusion. Uniting fracture faces could be met with steric obstacles if the repair device must go between the surfaces to repair them, since being between the two surfaces will tend to keep the surfaces apart and thus unrepaired.

But how are fracture surfaces likely to appear? Fractured surfaces will generally be cross-sections through various membrane-limited compartments (cells, myelinated axons, organelles), and planar separations between membrane bilayers. Within membrane-enclosed compartments, filamentous structures and molecular clusters such as enzyme complexes will be cleaved. In most cases, relatively free molecules such as cytoplasmic globular proteins should not be fractured, and the few that might be lost in this way can be neglected. Fractured microtubules, actin, etc. can be healed enzymatically. Steric hindrance is not a likely problem for individual molecules. Disrupted enzyme clusters can be reclustered (and will often recluster spontaneously when warmed sufficiently [47]).

In the case of membrane-bounded compartments that have been cleaved by fractures, one strategy would be to heal the limiting membrane first. It will not be destabilizing to remove replica material from membranes fractured perpendicular to the plane of the membrane because membranes can be adequately stabilized from above and below the plane of the membrane before the replica material is removed. As the naked membrane faces are brought together, they will tend to fuse spontaneously (48). This is also true for bringing together membranes fractured between leaflets. No specific chemical bonding will have to be induced to heal the major portion of the fracture.

If membrane fluidity is too low to permit good fusion at the prevailing temperatures, a small amount of specialty lipid can be added to the local area to enhance fluidity sufficiently to permit fusion to occur (49). After membrane fusion has occurred, some individual molecular species (particularly cross-linked proteins) associated with the formerly fractured area of the membrane might exist in a damaged (cleaved) form. These damaged components can be examined later, at higher temperatures, where they can be enzymatically healed (50).

The result would be a resealed compartment containing an internal plate of replica material. This material can then be disassembled from the plate molecule by molecule. As structures are uncovered by this process that require covalent bonding, they can be rotated for access, bonded, and then rotated back into proper position as healing proceeds. As healing proceeds, the liberated replica material can be passed through the healed membrane and exported to the extracellular space for removal by conveyors to outside the brain.

Some fractures are bound to penetrate through debris fields resulting from axon tearing or from myelin unravelling. The actual fracture healing in such areas should be relatively trivial, since there is no organized structure on either side of the gap that must be reconstructed. The area will consist mostly of evacuated empty space (now filled with inert gas and/or water vapor), from which removal of replica material should be particularly easy. Since all debris have been previously immobilized, repair of fractures through the debris will not endanger the information content of the region.

Note that it is the cells, vascular bed, and extracellular scaffolding whose fracture faces should be healed first. The presence of fracture replicas in gas pockets that previously consisted of extracellular ice or glass is important for maintaining the registry of cell surfaces and should be maintained until cell surfaces are safely healed.

It must be recognized that even at -80 degrees C, most relevant chemical reactions, even with the benefit of customized enzymatic catalysis, will proceed very slowly if at all (22,45). The missing energy can be made up in a variety of ways. First, heat could be liberated highly locally to permit reactions to proceed. (It could be helpful in this regard that the extracellular space has been replaced with gas, which is a good thermal insulator.) Second, exotic chemicals (perhaps including customized free radicals [51]) could be used to do the covalent bonding necessary to heal individual fractured molecules. Although this would most likely result, in most cases, in molecules containing unnatural structures, these foreign structures could be removed and corrected at higher temperatures at which the proper types of chemistry are feasible. Finally, the option exists, if all others fail, to simply hold fractured molecules together with molecular clamps until such time as they can be chemically bonded at higher temperatures.

How much time is available for these manipulations at about -80 degrees C. Although it is not possible to be certain, the normal rule of thumb would be that several months of storage at this temperature should be possible without any appreciable intracellular deterioration (52). This should be more than enough time to carry out the required extracellular repairs.

VIII. Cell Repair

1. Debris consolidation-- Having healed the fractures at about -80 degrees C, the next major extracellular task is to redistribute cellular debris to their proper locations. The actual transport is simplified by the absence of extracellular diffusive barriers. Reinsertion of lipids and proteins into membranes and into cytoplasm proceeds by means of specialized transport devices, which could be individually powered by reactive chemicals supplied continuously on the molecular conveyor system from outside. Once repositioned, lipids will remain positioned through ordinary self-assembly mechanisms (given an aqueous intracellular phase and a thin layer of aqueous extracellular fluid).

Having previously mapped and analyzed all debris down to the molecular level, actual reconstruction of debris into tissue should be relatively straightforward. Intracellular proteins, once deposited in the proper sites, can be covalently bound in position or immobilized with molecular clamps. To facilitate self- assembly, the temperature may be raised to perhaps -60 degrees C for up to a few weeks (52,53) if need be, either early, late, or intermittently during the reassembly process. In cases in which debris are the result of extrusion of material from contracting membranes, their redistribution is delayed pending cell volume re-expansion. At this stage, we repair only debris resulting from tearing and the like.


2. Stabilization against diffusional/biochemical deterioration. While limited, some diffusion is possible in cytoplasm at -60 degrees C. We exploit this by introducing metabolic inhibitors at this temperature into the cytoplasm. Since we have ready access to the external surfaces of cells, we can easily deposit inhibitors at regular intervals along cell processes. The inhibitors are designed to block the action of any enzymes that permit catabolism to proceed to beyond an acceptable point. Once deposited, they can be ignored, since these relatively low molecular weight inhibitors will reach their targets by diffusion as rapidly as the normal substrates would otherwise reach these catabolic enzymes.

With the possibility of detrimental catalyzed chemical change precluded, the only further types of damage are diffusion (e.g., organelle swelling), spontaneous chemical modifications (e.g., oxidation, racemization, etc.), and structural collapse (due to declining cellular rigidity with rising temperature, causing cellular structures to sag in the absence of extracellular supports). We can ignore diffusional change at this stage because cells and organelles are all highly shrunken. Random chemical damage can be ignored at this point and will be addressed later. Minor modifications to the extracellular communications network, which can double as a kind of extracellular "connective tissue," are now made to ensure the prevention of sagging during continued warming.

In order to further prepare for warming, nanocomputer-based cell repair machines similar to those described by Drexler (1) are introduced into the cytoplasm at this time. Although they are incapable of effecting rapid repairs at -60 degrees C, their introduction at this time allows them to begin repairs at the first opportunity during warming. They may well be capable of carrying out extensive sensing and computational functions at -60 degrees C in preparation for their actual repair activities at higher temperatures.


3. Cell Volume Restoration. As noted in the discussion on mechanisms of damage, it is during thawing that many problems arise. In the present scenario, no ice is present anywhere throughout the brain and, thus, the brain never has to go through the process of thawing per se. We do,
however, ultimately have to return cell volumes and cell water contents to normal. Our advantage is that we can do this in whatever manner is most desirable: we can expand cell volume at temperatures lower than would normally be associated with volume expansion during thawing (by adding both glycerol and water to the cells, we could fully expand cell volume even at -60 degrees C if we so chose), or we can expand cell volume at higher temperatures than would occur during thawing (by failing to add water to the same extent as it would be supplied by the progressive melting of ice).

The assumption we will make here is that we wish to do the former: expand the cells at temperatures lower than would exist during thawing. The reasoning is that there are many types of cellular injury which ultimately must be dealt with. If we rehydrate in a manner that simulates normal thawing, we tend to have to deal with all of these problems simultaneously. By re-expanding our cells at temperatures in the vicinity of -60 to -30 degrees C, rather than in the normal range of thawing (11) (i.e., -40 to -8 degrees C), we can take care of membrane re-expansion issues more-or-less independently of metabolic issues. We may want to favor the highest temperatures for re-expansion that do not begin to induce appreciable metabolic problems so that we can maximize membrane fluidity and minimize problems that may arise from unreversed membrane lipid phase transitions during cellular and membrane re-expansion.

Before cell re-expansion can proceed, there must be sufficient extracellular volume available for the re-expansion. We thus withdraw a portion of the extracellular communications network, much of which has already accomplished its purposes and is no longer needed. We leave in place conveyors for water and for glycerol, cellular supports, and assorted other devices.

We thus begin, at about -60 degrees C, to transport glycerol and water into the cytosol and axoplasm so as to maintain a ratio of glycerol to water that has a freezing point of about -61 degrees C. This process is carefully coordinated with the process of re-inserting extruded membrane material. As these two processes proceed, we also gradually raise the temperature, adjusting our membrane transporters to convey more and more water in comparison to glycerol so as to maintain an intracellular freezing point just below the prevailing temperature. Transport could again be powered by highly reactive chemicals introduced by conveyors.

At -30 degrees C or so, most (but not all) lost volume and all formerly extruded membrane material has been replaced. (We retain some extracellular space for the continued presence of some supporting devices.) The extracellular machinery for processing extruded material is withdrawn. The cells contain more than 6 M glycerol, a higher concentration than they began with. This is a sufficient concentration to preclude most intracellular chemistry, particularly at the prevailing temperature. The metabolic inhibitors introduced earlier have diffused to their targets and inactivated them. While cell volume expansion has proceeded, similar volume control measures have been completed for intracellular organelles. These measures have automatically included reversal of pre-existing organelle swelling. Other membrane transporters have also had sufficient opportunity to reverse ionic (Na+, K+, Ca+*, C1-, etc.) imbalances in both the cytoplasm and in organelles. They will continue to be active until brain temperature is returned to near normal values.

Volume control measures will not be entirely successful unless significant membrane phase changes have been reversed by this point in the repair process. This may happen spontaneously due to the elevation of temperature but, if not, it will be induced at this time by the temporary insertion of specialty lipid or molecules such as trehalose (54) (most likely in combination with more direct means).


4. Rehydration. At this point, the extracellular space can be flooded with glycerol-water-salt-substrate-colloid solution. This is done to maintain membrane integrity and to simplify water transport during rehydration. Colloid will preclude cell swelling in the cold without the need for vigorous ionic pumping (55).

We now reverse the direction of the membrane glycerol transporters and slowly transport glycerol to the extracellular space at the same time the glycerol concentration in the extracellular space is similarly being reduced by transport to the outside. By equating the proportion of cell glycerol removed to the proportion of extracellular glycerol removed, water activity is kept identical in the two compartments without a change in cell volume due to spontaneous diffusion of extracellular water into the cells. (Water diffusion should be sufficiently fast to preclude the need for specific -- and very energy intensive transmembrane water pumping at this stage.) At all times, the glycerol concentration within the cells is just sufficient to prevent the cytosol from freezing. Eventually, we arrive at 0 degrees C and a glycerol concentration of about 150mM. More of the extracellular communications and conveyance system is withdrawn.

Having reconstructed cellular and extracellular structures on a gross level, the vascular system is now sufficiently intact to permit cerebral perfusion to be reinstated. The brain vasculature should remain intact for days at 0 degrees C even without extensive protective modifications provided it has been sufficiently well repaired (56). The perfusate contains necessary substrates, repair machines, and psychrophillic anabolocytes, as required. Given that organisms have been found in nature that can grow at temperatures as low as 20 degrees C below zero (57), vigorous repairs are clearly possible at O degrees C, despite the very low metabolic rate of the original tissue.

These new devices as well as the previously-deployed intracellular cell repair machines therefore now proceed to correct the most critical types of continuing damage. Their activities may include, for example: myelin synthesis and replacement; bacterial and viral killing; protein reaggregation; cytoskeletal reassembly; reversal of glycerol-induced biochemical reactions; de novo synthesis of key missing proteins and other key metabolites; reversal of exotic, unnatural chemical bonds formed in order to heal otherwise intractable lesions at lower temperatures; removal of specialty lipid; restoration of normal intracellular buffering and pH; repair or removal of peroxidized, racemized, oxidized, or otherwise modified structures, resulting in their replacement with undamaged structures. Repair is powered by the chemical energy stored in the remaining glycerol present in the cells (precluding the need to otherwise dispose of this glycerol and completing the return to isotonicity) as well as by special chemical energy sources now available from the perfusate.

Almost all of the extracellular communications and general support network is now disassembled and withdrawn.

IX. Metabolic Restoration

As temperature is elevated further, oxygen is reintroduced and many metabolic inhibitors are degraded or inactivated. Necessary protein repairs are completed. Successful repair is checked by examining certain key metabolite behaviors in each significant metabolic pathway that are indicative of proper metabolic startup. Departures from expectation are diagnostic of any lingering underlying problems, which are then specifically corrected to the degree necessary. The required fine-tuning adjustments could be carried out largely by de novo synthesis of deficient proteins, by supplying inhibitory metabolites that are normally present and needed to control the overactivity of other proteins, by providing necessary protein cofactors that were previously lost, etc.

The synthesis of purely artificial proteins required for specialty tasks may also be called for. Protein denaturation is reversed artificially at this point as may be needed. [Renaturation could be accomplished by a variety of techniques. For example: a) The protein could be completely unfolded by seizing it at the N- or C-terminal end and passing it as a straight chain through a molecular tunnel resembling the channel nascent polypeptide chains pass through as they emerge from ribosomes, then allowing the emerging protein to refold either spontaneously or in cooperation with existing intracellular folding "chaperones" (58); or b) the protein could be attached to a scaffolding whose shape is altered in such a way as to renature the protein or allow the protein to complete spontaneous renaturation when released from the scaffolding after shape alteration.] After these diagnostics and fine-tuning tasks are completed, metabolism is released from artificial control.

X. Disease Reversal and Reanimation

Brain temperature is raised to 25-37 degrees C. Cell metabolism may still be grossly abnormal in a variety of ways: it will not have been necessary to have previously reversed all details of the previously existing pathological state, but only those details required for subsequent cellular self-maintenance and self-repair. Cells "know" what their proper state is and will spontaneously establish that state provided they are viable enough to continue to exist and to repair themselves.

While this is happening, conventional medical nanotechnology will be at work on specific disease processes, reversing them, establishing proper connections to extracephalic structures, and, if need be, assisting with the provision of a new body. Given stable physiology, these curative processes, including the partial or even complete reversal of aging, can be allowed to proceed as long as needed. Very few constraints on repair exist at this point. Technologies for dealing with specific disease states will be routine and powerful and require no description here.

Once the patient has been restored to a state approaching perfect physical health, consciousness is restored.


Summary and Conclusion

A "realistic" scenario for the repair of the frozen brain is proposed. It is based on the specific details of freezing injury and on the natural resistance of most cellular constituents to freezing damage, as well as on the natural self-assembly and self-repair of living cells. It avoids the need for performing chemical reactions below the glass transition temperature while at the same time avoiding the problems of diffusive information loss on warming. Although each step has not yet been subjected to thorough analysis, each is concrete and based on known fact. The scenario is fully open to criticism, testing, and refinement. It thus could serve as a basis for future discussions of the feasibility of moderate approaches to the restoration of those frozen by today's technology.


Disclaimer

This scenario is predicated on many assumptions--such as the assumption of adequate preservation by current technology-that may be false. This scenario does not prove that cryonics can or will succeed. It may, however, facilitate discussion of that possibility.


References and Notes

1. Drexler, Eric, Engines of Creation, Anchor/Doubleday, New York, 1986.

2. Drexler, Eric, Cell Repair Machines and Tissue Reconstruction: Some Notes on Computational Complexity and Physical Constraints, (unpublished document, 1/12/1984, pp. 1-53).

3. Merkle, Ralph, Molecular Repair of the Brain, Cryonics, 10(10), 21-44 (Oct, 1989).

4. Fahy, G.M., Molecular Repair Of The Brain: A Scientific Critique, Cryonics, 12(2), 8-11(Feb, 1991).

5. Leaf, J.D., M.G. Federowicz, H. Hixon, Appendix A: The "Smith Criterion" for adequate cryoprotection, Cryonics, 6(11), 38 (Nov, 1985).

6. Fahy, G.M., T. Takahashi, and A.M. Crane, Histological cryoprotection of rat and rabbit brains, Cryo-Letters, 5, 33-46 (1984),

7. Fahy, G.M., T. Takahashi, A.M. Crane, and L. Sokoloff, Cryoprotection of the mammalian brain, Cryobiology, 18, 618 (1981); Fahy, G.M., and A.M. Crane, Histological cryoprotection of rabbit brain with 3M glycerol, Cryobiology, 21, 704, (1984).

8. Burch, H.B., O.H. Lowry, et al, Effect of fructose, dihydroxyacetone, glycerol, and glucose on metabolites and related compounds in liver and kidney, J. Biol. Chem., 245, 2092-2102 (1970).

9. Tanizaki, M.M., H.M.S. Bittencourt, et al, Activation of low molecular weight acid phosphatase from bovine brain by purines and glycerol, Biochem. Biophys. Acta, 485, 116-123 (1977).

10. Bello, J., The stare of tyrosines of bovine pancreatic ribonuclease in ethylene glycol and glycerol, Biochemistry, 8, 4535-4541

11. Jochem, M., and Ch. Korber, Extended phase diagram for the ternary solutions H2O-NaCl-glycerol and H2O-NaCl-hydroxyethyI starch (HES) determined by DSC, Cryobiology, 24, 513-36(1987).

12. Mazur, P., Freezing of living cells: mechanisms and implications, Ant. J. Physiol., 247 (Cell Physiol., 16), C125 (1984).

13. Steponkus, P.L., M.F. Dowgert, et al, Destabilization of isolated plant protoplasts during a freeze-thaw cycle: the influence of cold acclimation, Cryobiology, 20, 448-65 (1983).

14. Steponkus, P.L., D.V. Lynch, et al, Plant cryobiology: cellular and molecular aspects of freezing injury and cold acclimation, in Low Temperature Biotechnology: Emerging Applications and Engineering Contributions, J.J. McGrath and K.R. Diller, Eds., ASME, New York, 1988, pp.47-56.

15. Williams, R.J., H.J. Hope, et al, Membrane collapse as a cause of osmotic injury and its reversibility in a hardy wheat, Cryobiology, 12, 554-5 (1975).

16. Fagg, G.E., E.E. Mena, et al, Freezing eliminates a specific population of L-glutamate receptors in synaptic membranes, Neurosci. Lett., 38, 157-62 (1983); Wu, R., R. Carlin, et al, Binding Of L-[H3]glutamate to fresh or frozen synaptic membrane and post-synaptic density fractions isolated from cerebral cortex and cerebellum of fresh or frozen canine brain, J. Neurochem, 46, 831-41 (1986). Protein loss has also been documented in other cells. For example, see Takahashi, T., S. Inada, et al, Osmotic stress and the freeze-thaw cycle cause shedding of Fc and C3b receptors by human polymorphonuclear leukocytes, J. Immunol., 134, 4062 (1985).

17. See, for example, Daniel L. Alkon's lay paper, Memory Storage and Neural Systems, Scientific American, pp. 42-50 (July, 1989), as well as his papers in Science (226, 1037-1045, 1984; 239, 998-1005, 1988), and Alkon's book, Memory Traces in the Brain, Cambridge University Press, 1988.

18. The Cryobiological Case for Cryonics, booklet available from the Alcor Life Extension Foundation; 7895 E. Acoma Dr., Scottsdale, AZ 85260. Tel: (800) 367-2228.

19. Hardy, J.A., P.R. Dodd, A.E. Oakley, R.H. Ferry, J.A. Edwardson, and A.M. Kidd, Metabolically active synaptosomes can be prepared from frozen rat and human brain, J. Neurochem., 40, 608-14 (1983).

20. Stahl, W.L., and P.D. Swanson, Effects of freezing and storage on subcellular fractionation of guinea pig and human brain, Neurobiology, 5, 393-400 (1975).

21. Kandel, E. R., and Schwartz, J. H., Principles of Neural Science, Second Edition. Elsevier, New York, 1985.

22. Franks, Felix, Biophysics and Biochemistry at Low Temperature, Cambridge University Press, New York, 1985

23. Alberts, B. D. Bray, et al, Molecular Biology of the Cell, 2nd Ed., Garland Publishing, Inc., New York, 1989.

24. Goodin, R., and J. Levitt, The cryoaggregation of bovine serum albumin, Cryobiology, 6, 333-8 (1970).

25. Lovelock, J.E., The mechanism of the cryoprotective action of glycerol against hemolysis by freezing and thawing, Biochem. Biophys. Acta, 11, 28-36 (1953).

26. Cryonics: Reaching for Tomorrow. Booklet available from Alcor Life Extension Foundation.

27. Hardy, J.A., P.R. Dodd, A.E. Oakley, R.H. Ferry, J.A. Edwardson, and A.M. Kidd, Metabolically active synaptosomes can be prepared from frozen rat and human brain, J. Neurochem, 40, 608-14 (1983).

28. Pegg, D.E., and M.P. Diaper, On the mechanism of the protective action of glycerol, Biophysical Journal, 54, 471-88 (1988).

29. Houle, J.D. and G.D. Das, Cryopreservation of embryonic neural tissue and its successful transplantation in the rat brain, Anat. Rec., 196, 81A (1980); Houle, J.D. and G.D. Das, Freezing of embryonic neural tissue and its transplantation in the rat brain, Brain Res., 192, 570-4 (1980); Houle, J.D. and G.D. Das, Freezing and transplantation of brain tissue in rats, Experientia, 36, 1114-5 (1980); Das, G.D., J.D. Houle, J. Brasko, and K.G. Das, Freezing of neural tissues and their transplantation in the brain of rats: technical details and histological observations, J. Neurosci. Methods, 8, 1-15 (1983); Jensen, S., T. Sorenson, A.G. Moller, and J. Zimmer, Intraocular grafts of fresh and freeze-stored rat hippocampal tissue: a comparison of survivability and histological and connective organization, J. Comp. Neurol., 227, 558-68 (1984); Sorenson, T., S. Jensen, A.G. Moiler, and J. Zimmer, Intracephalic transplants of freeze-stored rat hippocampal tissue, J. Comp. Neurol., 252, 468-82 (1986); Jensen, 8., T. Sorenson, and J. Zimmer, Cryopreservation of fetal rat brain tissue later used for intracerebral transplantation, Cryobiology, 24, 120-34 (1987)

30. Suda, I, K. Kite, and C. Adachi, Bioelectric discharges of isolated cat brain after revival from years of frozen storage, Brain Res., 70, 527-31 (1974).

31. Baker, P.F., A.L. Hodgkin, et al, Replacement of the axoplasm of giant nerve fibers with artificial solutions, J. Physiol. (London), 164, 330-54 (1962).

32. Kylin, H., Uber die Kalteresistenze der Meeresalgen, Eer. dtsch. bor. Ges., 35, 370-84 (1917); van den Berg, L., and P.S. Soliman, Effect of glycerol and dimethyl sulfoxide on changes in composition and pH of buffer salt solutions during freezing, Cryobiology, 6, 93-7 (1969).

33. Persidsky, M.D., Lysosomes as primary targets of cryoinjury, Cryobiology, 8, 482-8 (1971); Persidsky, M.D., and Ellett, M.H., Lysosomes and cell cryoinjury, Cryobiology, 8, 345-9 (1971).

34. Steponkus, P.L., and D.V. Lynch, Freeze/thaw-induced destabilization of the plasma membrane and the effects of cold acclimation, J. Bioenergetics Biomembranes, 21, 21-41(1989).

35. Koyova, R.D., B.G. Tenchov, et al, Sugars favor formation of hexagonal (HII) phase at the expense of lamellar liquid-crystalline phase in hydrated phosphatidylethanolamines, Biochem. Biophys. Acta, 980, 377-80 (1989).

36. Lyons, J.M., Phase transitions and control of cellular metabolism at low temperatures, Cryobiology, 9, 341-50 (1972).

37. Quinn, P.J., A lipid-phase separation model of low temperature damage to biological membranes, Cryobiology, 22, 128-46(1985).

38. Jacobsen, I.A., Pegg, D.E., et al, Introduction and removal of cryoprotective agents with rabbit kidneys: Assessment by transplantation, Cryobiology, 25, 285-99, (1988).

39. Fahy, G.M., Analysis of "solution effects" injury: Cooling rate dependence of the functional and morphological sequellae of freezing in rabbit renal cortex protected with dimethyl sulfonide, Cryobiology, 18, 550-70 (1981).

40. Unpublished experimental results of Alcor Life Extension Foundation and Cryovita Laboratories.

41. Fahy, G.M., J. Saur, et al, Physical problems with the vitrification of large biological systems, Cryobiology, 27, 492-510 (1990).

42. Gadea-Ciria, M., J. Gervas Camacho, et al, Water content of various regions of the feline nervous system, Medical Biol., 53, 469-74 (1975).

43. J.G. Linner and S.A. Livesey give the following calculations in their chapter, Low Temperature Molecular Distillation Drying of Cryofixed Biological Samples, in Low Temperature Biotechnology, Emerging Applications and Engineering Contributions, J.J. McGrath and K.R. Diller, Eds., ASME, New York,1988,pp.147-157.

44. Orci, L., and A. Parrelet, Freeze-Etch Histology. A Comparison Between Thin Sections and Freeze-Etch Replicas, Springer-Verlag, New York, 1973.

45. Douzou, P., Cryobiochemistry, An introduction, Academic Press, New York, 1977.

46. Kroener C., and B. Luyet, Formation of Cracks During the Vitrification of Glycerol Solutions and Disappearance of the Cracks During Rewarming, Biodynamica, 10, 47-52 (1966).

47. Both proteins and membranes exist in the form they do because of the immiscibility of water and hydrocarbons. This immiscibility causes these structures to self-assemble spontaneously if permitted to do so; this is the basis of spontaneous protein renaturation and membrane assembly. Protein clusters often involve hydrophobic associations as well, but even when other contributions are more important, mis-clustering is, in principle, equally spontaneously reversible. Self-assembly can happen incorrectly, but, given a little guidance, can surely be directed to happen correctly. See also note 48 below.

48. Fat exposed to water preferentially associates with other fat if any is available. A fractured membrane presents two "greasy" surfaces to water, which is entropically unfavorable; it is thermodynamically favorable for these two cut surfaces to fuse together so as to eliminate the unfavorable water-fat interface. This tendency is, however, reduced by low temperatures (which reduce the energy cost of hydrating fat) and by solidification of the membrane. A good general discussion of these issues is given in The Hydrophobic Effect, Formation of Micelles and Biological Membranes, by Charles Tanford (2nd Edition, 1980, John Wiley & Sons, New York). As Tanford notes (p. vii): "The hydrophobic effect is perhaps the most important single factor in the organization of the constituent molecules of living matter into complex structural entities such as cell membranes and organelles."

49. "Specialty lipids" could be made by reducing the number of carbon atoms in the fatty acid tails of ordinary membrane phospholipids, increasing the number of double bonds (especially cis double bonds) in these tails, fluorinating the fatty acid tails, modifying polar head groups to prevent close association of the lipid tails (by preventing clumping of these head groups), or by any combination of these maneuvers. These modifications are all known to reduce the freezing points of lipids and/or hydrocarbons and hence increase their fluidity. (See also: Small, D. M., et al, The Physical Chemistry of lipids: From Alkanes to Phospholipids. Plenum Press, New York, 1986 [Handbook of Lipid Research, Vol. 4].) Specialty lipids based on such modifications should, therefore, enhance the ability of lipid phases doped with them to fuse. It does not seem likely that the specialty lipids must reverse membrane phase separations to effect membrane fusion. Even small, free molecules such as pentane or its relatives could suffice: as long as the molecule is insoluble in water and preferentially associates with hydrophobic species, it should produce the desired effect. (The fact that membranes continue to exist at low temperatures suggests that hydrophobic forces will remain strong enough at these temperatures to promote self-assembly of hydrophobic entities in an aqueous environment.) Even if a molecule such as pentane becomes volatile on warming, the membrane will not be affected, provided it becomes sufficiently fluid before the small species evaporates.

50. Evidence that it is permissible to heal some fracture damage at higher temperatures is provided by the results of Dr. Luiz de Medinaceli, who found he could regenerate rat sciatic nerves that he had first frozen and then cut cleanly at temperatures just below 0 degrees C. The nerve ends were held together by special tethers and extracellular potassium was elevated to keep the cut axons alive. Only very cleanly cut nerves regenerated well. His work was discussed in a series of papers that appeared in Experimental Neurology, Volume 81 (pages 459-468; 469-487; and 488-496) and Volume 84 (396-408), in 1983 and 1984. See particularly Vol. 81, pp. 469-496. His work is now being extended to human patients (personal communication).

51. As Mazur discusses in reference 12, free radical reactions can proceed relatively unabated regardless of temperature, owing to the lack of any activation energy for these reactions.

52. Meryman, H.T., Review of biological freezing, in Cryobiology, H.T. Meryman, Ed., Academic Press, New York, 1966, pp 1-114.

53. Suda's papers suggest that brains will be stable at such temperatures for at least this long, and probably for much longer. See reference 30 and the following reference: Suda, I., K. Kite, and C. Adachi, Viability of long-term frozen cat brain in vitro, Nature, 212, 268-70 (1966).

54. Crowe, J.H. and L.M. Crowe, Interaction of sugars with membranes, Biochem. Biophys. Acta, 947, 367-84 (1988).

55. Hitchcock, D.I., Proteins and the Donnan equilibrium, Physiol. Rev., 4, 505-531 (1924); Leaf, A., Regulation of intracellular fluid volume and disease, Am. J. Med., 49, 291-5 (1970); Mendler, N., H.J. Reulen, et al, Cold swelling and energy metabolism in the hypothermic brain of rats and dogs, in Hibernation and Hypothermia, Perspectives and Challenges, F.E. South, J.P. Hannon, et al, Eds., Elsevier, New York, 1972, pp. 167-190.

56. White, R.J., M.S. Albin, et al, Prolonged whole brain refrigeration with electrical and metabolic recovery, Nature, 209, 1320 (1966).

57. Actual growth has been confirmed at -12 degrees C, and unconfirmed reports of cell growth at -18 degrees C, -20 degrees C, and -34 degrees C are available: see Mazur, P., Limits to life at low temperatures and at reduced water contents and water activities, Origin of Life, 10, 137-59 (1980). Continuing metabolism has been documented to occur at -30 to -40 degrees C by S.M. Siegel, T. Speitel, et al, Life in Earth extreme environments: a study of cryo-biotic potentialities, Cryobiology, 6, 160-81 (1969).

58. For some recent references, see H. Blumberg and P.A. Silver, A homologue of the bacterial heat-shock gene DnaJ that alters protein sorting in yeast, Nature, 349, 627-30 (1991).