Summary

After blood flow to the brain stops, the gel-like networks that make up the brain begin to decay. Functional properties are lost quicker than structural ones. Biomolecular and structural decomposition is not immediate but is a delayed process that occurs on the timescale of minutes to hours to days. The speed of decomposition depends on the environmental conditions, most notably the temperature of the brain. While a long enough post-mortem interval will eventually lead to information-theoretic death, there exists a window of unknown duration after circulation to the brain stops but before information-theoretic death occurs during which it makes sense to preserve the brain.

Our review article

These notes draw upon a recent collaborative review paper on postmortem changes in brain cell structure that I was a part of (Krassner et al., 2023). If you’re interested in this topic, I recommend you read it. To ensure the recognition of my colleagues’ work, I want to make sure to differentiate my personal notes from this joint review paper. These notes are my personal thoughts, related to brain preservation as opposed to brain banking, and focus more on macroscopic, functional, and biomolecular postmortem changes, which are not as fully covered in our review paper.

What’s the point in preserving “dead” people?

When people first hear of cryonics or brain preservation, many wonder what is the point or how it could be possible if it is done on “dead” people. First, to answer the question directly, it is important to distinguish between legal death and information-theoretic death. Death is a process. There is good reason to think that people we declare legally dead on the basis of our medical technology today might not be declared as legally dead based on future standards and technology. Thus, it is reasonable and in an important sense conservative to preserve the brain for those who desire this in case there is a chance that they are able to be revived if future technology develops to make this feasible and humane. This is discussed in a previous chapter.

Second, this discussion is in part forced upon those interested in brain preservation by paternalistic and unjust medicolegal systems across the world that do not allow people to exercise autonomy over their bodies (in many ways; this is just one). Ideally, people could begin the procedure at the time of their legal death in the case of a terminal condition if that is their preference. The discussion can therefore become circular, with some people saying “it won’t work because they are already dead” and others (or sometimes even the same people) saying “you’re not allowed to do it at the time of legal death because it won’t work”. I use the word “systems” here intentionally because the rules and regulations are usually outside of any one individual’s control.

However, within the confines of society’s current laws, in this chapter I address the question of whether it is worthwhile to perform brain preservation at all if one must wait until after legal death. The evidence suggests the answer is clearly yes. And even if society does one day offer people the autonomy to choose the time and manner of their deaths in the case of having a terminal condition, this discussion will still be relevant if someone legally dies in an unexpected manner. For these reasons, the postmortem interval is an enormously important metric in brain preservation. In an ideal world, it would be understood, measured, and minimized.

Global cerebral ischemia as a final common outcome that initiates the post-mortem interval

With the exception of brain death and a few other very rare causes of death, the ultimate final outcome in legal death is the failure of the cardiovascular system and a resulting lack of blood flow to the brain. Without oxygen and nutrients supplied by the constant flood of blow, the proteins maintaining the function and shape of brain cells begin to break down. This is generally the type of damage that people are referring to when they speak of the decomposition that occurs during the post-mortem interval prior to preservation.

The term used for the lack of blood flow to the brain is called “global cerebral ischemia.” Global refers to the whole brain rather than a particular area. Cerebral obviously refers to the brain. And ischemia is the technical term for lack of blood flow.

Global cerebral ischemia is fairly well studied because it is relevant to brain function following recovery from cardiac arrest. There is a concept called the ischemic cascade that is thought to characterize the sequence of stereotyped biochemical changes that tend to occur in ischemic tissue such as the brain.

However, most of the research on global cerebral ischemia focuses on the effects of short-term global cerebral ischemia and the long-term sequelae when blood flow returns, which is called reperfusion.

When learning about decomposition of brain tissue, it is common to hear that the brain can be severely damaged by 5-10 minutes or more minutes of lack of blood flow to the brain at room temperature. This is true, although the damage does not occur immediately. Short periods of ischemia lead to delayed tissue damage mechanisms that occur after the re-initiation of blood flow to the brain, which is called reperfusion injury.

The most prominent aspect of reperfusion injury is delayed cell death triggered by biochemical signals that accumulate during the short-term ischemia (Lipton, 1999). Delayed cell death seems to occur due to mechanisms that “tag” brain cells exposed to brief ischemia for subsequent destruction, for example via the expression of cytokines, over the timescale of 6 hours to several days.

For example, 10-30 minutes of reversible lack of blood flow to the brain eventually leads to significant cell death in the hippocampus, but data suggests this damage cannot be seen microscopically for up to 24 hours following restoration of blood flow (as cited in (Pulsinelli, 1985)). Other data suggests that neuronal loss is not detectable at 48 hours after a brief period of global cerebral ischemia, but is detectable after 7 days (as cited in (Font-Belmonte et al., 2019)).

Consistent with the idea that the most damaging aspects of global cerebral ischemia is reperfusion, studies have shown that by regulating aspects of reperfusion, it is possible to substantially mitigate damage to the brain tissue (Lindblom et al., 2021).

From the perspective of brain preservation, there are a few important take-aways from this finding that much of the damage associated with global cerebral ischemia is delayed:

The first is that we should be judicious about what is included in the perfusate for any perfusion-based preservation methods. Including glucose and oxygen in an attempt to mitigate biochemical changes, while possible (Vrselja et al., 2019), might actually cause more brain tissue decomposition. It may be better to let sleeping dogs lie.

Second, using brain tissue damage following a certain number of minutes of lack of blood flow to the brain as a proxy for the amount of postmortem damage after an equivalent amount of minutes is much more pessimistic about timelines than is called for from the perspective of brain preservation. We can avoid reperfusion injury because we don’t plan to perform reperfusion with oxygen or nutrients until revival. Preventing reperfusion injury is just another problem to be solved or routed around prior to revival, which seems relatively trivial compared to the other massive problems that will have to be solved.

Third, much of the literature on the topic of global cerebral ischemia is about active decomposition that requires restoration of blood flow to allow for coordinated functions that leads to subsequent programmed cell death. Most active processes will not be present during an extended postmortem interval, as the ATP pools will be rapidly depleted. We are instead interested in the more focused literature that speaks to passive decomposition during global cerebral ischemia without reperfusion.

Mechanisms of postmortem brain decomposition

The two major mechanisms of decomposition are autolysis and putrefaction (Teo et al., 2014).

Autolysis (“self” + “splitting”) is the process of cellular self-destruction. It is primarily performed by hydrolytic enzymes naturally produced by cells that eventually begin to break down the cell’s own biomolecules.

Putrefaction is due to degradation caused by microorganisms that are either present within the body at the time of death or migrate there afterward. Compared to other organs, the process of putrefaction is relatively more consistent in the brain its pattern of breakdown because it is more protected from microorganism invasion and environmental effects by the skull (M. Ith et al., 2002).

Of the two processes, autolysis usually occurs first and does not inflict as much damage on the brain, especially damage that can be observed by the naked eye. Putrefaction often begins in the brain at a post-mortem interval of approximately 1-3 days (Michael Ith et al., 2011). Because it usually happens more quickly, most of our focus will be on autolysis. Putrefaction might occur more quickly if there is a brain abscess, severe sepsis, or traumatic injury prior to legal death that introduces bacteria.

Change in gel-like properties of the brain after death

Without blood flow, the brain does not immediately turn into a liquid. The breakdown of cell cytoskeleton, extracellular matrix, and other structural components that bind the brain’s structure together requires time.

The framework that I currently favor is to think of brain cells and their extracellular environment as maintained via weak gel-like structures. Numerous pieces of evidence support this, including rapid gel-gel phase transition that can cause the formation of condensed “dark neurons” upon exposure to various types of structural damage (Kherani et al., 2008). In this framework, neuronal morphology is maintained by numerous types of biomolecules that, through their interactions, create and maintain gel-like structures.

As postmortem decomposition begins, the gel-like networks often aggregate into larger, artifactual structures. This is why aggregates are often seen in the nucleus in the early postmortem interval. For example, (Routtenberg et al., 1974) describe several types of aggregations during a 1-10 minute perfusion fixation delay of the brain:

The spherical densities observed in our post-mortem material may represent the aggregation of material in the cell cytoplasm. The impression gained by observation of our post-mortem material is of a cytoplasmic matrix (‘cytonet’) which aggregates or precipitates, increasing aggregation with increasing post-mortem time, and attaching to certain membranes: the outer nuclear envelope, endoplasmic reticulum and the post-synaptic specialization. This is similar to the theory of the coated vesicle artifact proposed by Gray (1973). If the densities represent aggregation of existing cytoplasmic material, then a parallel process can be observed in the nucleus where clumping of chromatin in post-mortem material is obvious. Perhaps not unrelated to these considerations is the aggregation of synaptic vesicles which is particularly pronounced in the 5 and 10 min post-mortem material.

Macroscopically, the brain actually has been been found to become more stiff immediately following death. One study found that brain stiffness increases around 2-fold 45 minutes after death, especially in white matter regions such as the corpus callosum (Weickenmeier et al., 2018). This is suggestive of a slight initial strengthening of gel-like structures, at least at a macroscopic scale.

Over the longer term, as decomposition continues, the brain will lose the microstructures that make it gel-like and it will eventually turn into a macroscopic liquid as it decomposes. For example, one study found that the brain’s storage modulus was reduced more than 3-fold after 24 hours, which was attributed to necrosis of the brain tissue (Tabet et al., 2019).

Another way to measure the breakdown of the actin cytoskeleton after death is to measure its gel-like properties, for example the Young’s modulus of neurons. One study found that the Young’s modulus of neurons was around 60 kpa at baseline (Mustata et al., 2010). It remained at around this level for 4 hours, after which it began to degrade and the elasticity therefore reduced. However, this study was in vitro, which based on other evidence likely meant there was a faster rate of decomposition than cells in situ.

Actin and cytoskeleton disassembly during global cerebral ischemia

There are many possible mechanisms for the disassembly of actin cytoskeleton networks. The most common causes of actin disassembly during most of one’s life tend to be active forms of depolymerization that require energy in the form of ATP. Because energy pools are lost soon after the lack of blood flow to the brain, actin cytoskeleton disassembly is more likely to be broken down by slower and passive severing and fragmentation mechanisms. There may even be protective factors against passive actin fragmentation during the post-mortem interval due to the presence of rods that bind to actin filaments and seem to stabilize their structure (Bernstein et al., 2006).

The major mechanism of the breakdown of actin networks and the cytoskeleton in general during global cerebral ischemia seems to be the dispersion of enzymes that break down proteins non-specifically. For example, as ATP is depleted, lysosome membranes can be disrupted and eventually rupture, causing the release of their catabolic enzymes into the cytosol. As a general rule, in cell death pathways, high levels of lysosome rupture is associated with downstream disruption of the cell membrane (Ono et al., 2003).

Lysosomes contain autolytic enzymes and postmortem decomposition is often attributed to lysosomal breakdown and associated release of autolytic enzymes. McKeown 1979 studied the role of lysosomes in rat brain postmortem autolysis (McKeown, 1979). In the postmortem period, some lysosomes seem to rupture, in what appeared to be an all-or-none effect. There also seems to be small increases in lysosomal membrane permeability. However, this study’s results suggest that lysosomes remain relatively intact during the first 6 hours at room temperature and that cooling the brain after this amount of time led to a decrease in lysosome decomposition. Based on this data, lysosomal breakdown does not seem to be an immediate phenomenon.

When cells lack the energy to maintain themselves, there is a series of biochemical changes that tend to occur that are described as necrosis. These changes include a loss of cell membrane integrity and a massive influx of calcium and water. The influx of calcium into cells can activate proteolytic enzymes that break down proteins, including actin filaments. It is still meaningful to speak of cells at this point because, during the process of necrosis, cell morphology tends to be maintained until close to the end. As one review notes “the overall configuration of the moribund cell [is] maintained until the very end in vivo” (Martin et al., 1998).

The fact that enzyme-driven autolysis is an important early contributor to brain decomposition is one of the reasons that the temperature is so important during brain decomposition. Lower temperature lowers the activity of enzymes that break down the gel-like structures in the brain. Lower temperature also makes water more viscous, which slows down diffusion.

How long does it take for brain cells to fully decompose during global cerebral ischemia? There are no clear answers and this will differ across cells within the same brain and across brains. This is not a widely studied topic because it is not considered to be very clinically relevant, as long periods of global cerebral ischemia are clearly not survivable by contemporary medical standards.

Rigor mortis does not occur in brain tissue. However, the timescale of rigor mortis in muscles may give us a window into approximately how long it takes for actin to be broken down after legal death in the absence of ATP, which is necessary for normal actin disassembly mechanisms. Rigor mortis often completes around 8 to 10 hours after death and its breakdown, which seems to be dependent on passive actin disassembly, disappears around 36 hours after death (Shrestha et al., 2023).

In focal ischemia, where only one part of the brain loses blood flow, timescales are a little bit better defined. Advanced cellular decomposition can be defined as neurons that look shrunken, autophagocytotic, or apoptotic. One review suggests that in focal ischemia, significant cellular changes begin to occur at approximately 6 hours and increase substantially by 12-24 hours, at which point approximately 80% of neurons will have undergone enough decomposition to morphologically qualify them as “dead” (Lipton, 1999). While breakdown of morphology in focal ischemia is a useful proxy to consider, evidence suggests it is more damaging to the local brain tissue than global cerebral ischemia, in large part because the temperature is higher. As a result, a better way to measure the decomposition of brain cells after prolonged global ischemia is likely by analyzing empirical data from brain banking studies. This was performed in our review paper (Krassner et al., 2023).

Macroscopic morphology changes

Perhaps the most obvious changes in brain structure after legal death are those can be be seen with the naked eye, which are called gross morphologic changes.

Within the field of forensic neuropathology, there are ways to qualitatively score the gross morphology of the brain. One such measure scores brains with the least decomposition as those that appear grossly normal, are able to be extracted from the brain without fragmentation, and have clear internal anatomy (Hayman et al., 2017). As the brain decomposition scores increase, the surface and then the internal parts of the brain soften and then turn to a liquid. The internal anatomy, presumably aspects such as the grey-white matter distinction, becomes unclear and then no longer recognizable. It no longer becomes possible to manually extract the brain from the skull as a whole without causing visible damage to the brain tissue.

Using this classification scheme, Jarvis Hayman described the rate of decomposition of the brain in 13 cases with different estimated post-mortem intervals (Hayman, 2013). This may give a practical sense of how the brain tends to decompose macroscopically:

Case Number PMI Estimate Brain Decomposition Score Liquefied State Ability to Remove from Skull
1 10 days 4 Paste-like brain No, only in pieces
2 6.6 days 2 Softening Yes, maintained its shape, although was becoming fragile
3 2.5 days 1 Slight softening Yes, able to be removed without sulci/gyri separation
4 2.25 days 2 Softening Yes, but flabby when removed
5 5.5 days 2 Slight softening Yes, but slight softening on removal
6 1 day* 2 Soft with pasty consistent and gas bubbles present Yes
7 5 days 2 Very soft Yes, but fragmented easily on removal
8 5-9 days, with heater on 4 Thick, amorphous paste No, brain flowed out of the skull on removal
9 9-12 days 4 Very soft, becoming fluid No, completely disintegrated upon removal
10 5-6 days 3 Soft, fragmented when sectioned Yes, but slight softening and distortion on removal
11 10 days 3 Very soft, flattening of sulci Yes, but almost fragmented when removed
12 9 days 3 Soft, firm paste-like consistency Yes, but soft
13 12 days 4 Very soft, semi-liquid No, unable to be removed without fragmenting

The general sequence of events and timescales of visible brain morphology changes over the post-mortem interval is corroborated by other reports. For example, one classification scheme has the following timeline of gross putrefaction in 29 brains (Moriya et al., 2004):

Of course, the decomposition of bodies and brains after legal death is not uniform. It depends on many factors including the temperature, atmospheric oxygen, moisture, and acidity of the brain tissue.

While the general trend is for brain tissue to turn into a liquid paste over a timescale of days to one or two weeks, it is intriguing to note that this is not the case for all brains. From an archaeological perspective, the brain is actually the soft tissue that is most commonly found to be well-preserved through natural means over the very long term (Morton-Hayward et al., 2020). This may be due to the unique role of intermediate filaments in the brain in maintaining the structure over time. It is possible that some pre-mortem or post-mortem conditions may favor aggregation of intermediate filaments, thus dramatically strengthening portions of the cytoskeletal network and leading to long-term dry preservation over the timescale of centuries.

Some grave sites can have well preserved brains even in the absence of any brain preservation procedures. For example, one set of brains exhumed 80 years after death had remarkably good natural preservation, which was speculated to be saponification, due to the acidic properties of the soil in which the burial occurred (Etxeberria et al., 2020):

Two of the brains found naturally preserved at the mass grave of La Pedraja; Source: (Etxeberria et al., 2020)
Two of the brains found naturally preserved at the mass grave of La Pedraja; Source: (Etxeberria et al., 2020)

Just to be clear, it’s worth pointing out that while brains can be naturally preserved, even in this state they are highly degraded. For example, brain tissue in this state is generally reduced in volume to one-fifth the volume of normal brain tissue (Morton-Hayward et al., 2020). Histologically, it is almost certain that the biomolecule-annotated connectome is not able to be inferred from this tissue, so I very strongly doubt that any memories could be recovered from brain tissue naturally preserved in this manner.

Functional brain changes during global cerebral ischemia

As a general rule, following the loss of blood flow to an organ, function is lost sooner than structure (Madea, 1994):

The concept of “supravitality” in forensic medicine; Source: (Madea, 1994)
The concept of “supravitality” in forensic medicine; Source: (Madea, 1994)

From the perspective of information conservation, functional properties are not necessary to preserve. However, when functional properties of the brain are still present, it means that the structures underlying those functions could not have been irreparably damaged. In other words, functional preservation screens off the need to consider structural preservation.

The human brain is large and uses a substantial amount of energy, approximately 27% of the body’s total basal metabolic consumption (Boyer et al., 2018). In the absence of blood flow, glucose and ATP stores are dramatically depleted within minutes (Vrselja et al., 2019). As a result, active electrochemical and biochemical functions tend to stop quite quickly.

Historically, consensus opinion was quite pessimistic about the recovery of the human brain after blood flow stops (Bailey, 2019). We might call this the “5 minute rule” (Leonard et al., 1991). However, recent data has caused a reappraisal of this dogma. As it turns out, brain functions are not irretrievably lost within only 5 minutes after blood flow stops (Bailey, 2019).

Global electrophysiology activity

Following total lack of blood flow to the brain, a particular sequence of events tends to occur:

  1. First, there is electrical silencing. During global cerebral ischemia, the EEG begins to degrade within 1-5 seconds (Ravindran et al., 2019). Loss of consciousness occurs in approximately 4-10 seconds (Bailey, 2019). Lack of global electrical activity within the brain occurs within 20 to 40 seconds of global ischemia.

This rapid electrical silencing has been called an “austerity program.” It is thought to be an active process by which neurons recognize that their energy stores have been depleted and attempt to conserve them, in case energy can be restored. It may be produced by an oxygen sensor in brain cells that inhibits neuronal activity upon detecting a decline in the local oxygen levels. This phenomenon is called “nonspreading depression” because it tends to be a local phenomenon and has not been found to spread in the vertical or horizontal directions of the brain.

  1. The next step is called terminal spreading depolarizations, which are also called “brain tsunamis”. Within a few additional tens of seconds to a few minutes without blood flow, the electrical silencing austerity programs break down. There are then waves of electrical activity, called spreading depolarizations, that can be detected by electrodes within the brain (Dreier et al., 2018). This terminal spreading depolarization can be thought of as the last coordinated activity that one’s brain will perform in the absence of resuscitation. After the terminal spreading depolarizations occur, there is generally no more spontaneous electrical activity in the brain in the absence of intervention. It is considered electrically silent.

There appears to be no strict measure of when these terminal spreading depolarizations become irreversible (Dreier et al., 2018). As mentiond above, early lore, for example in the 1950s, was that brain function recovery was not possible if complete lack of blood flow was sustained for more than 5 minutes at room temperature (Siesjö, 1981). Experimental studies have since pushed this boundary substantially. For example, one study in monkeys have found that revival of monkeys with neurologic function intact could be pushed for up to 16 minutes as long as there was protection to the heart, which allows for rapid resumption of blood flow with adequate blood pressure to the brain once the intervention is reversed (Miller et al., 1970) (Miller et al., 1972).

  1. Cellular dysfunction leading to a late negative ultraslow potential. After the brain tsunamis stop, what can be measured is a slow decrease in the electrical potential of brain tissue. This likely occurs as a result of the breakdown of cell membrane ion channel and pump proteins. This is called the late negative ultraslow potential (Dreier et al., 2018).

Two major ways to measure global electrochemical activity in the brain are EEG and ECOG. EEG measures electricity on the surface of the skull, whereas ECOG measures electricity on the surface of the brain tissue.

Charpak 1998 found that the EEG was electrically silent in isolated guinea pig brains after 1 hour postmortem (Charpak et al., 1998). In this study, raising extracellular potassium concentrations and adding a GABA-A receptor antagonist to an isolated guinea pig brain that had undergone a PMI of 75 minutes was sufficient to cause spontaneous epileptiform bursts of electrical activity in the bilateral temporal lobes, although this is clearly not a physiological stimulation.

Vrselja 2019 found that ECOG was electrically silent after 4 hours postmortem without oxygen (Vrselja et al., 2019).

Two studies by Hossman and colleagues in the early 1970s showed that global electrophysiology (EEG or ECOG) could be restored after 1 hour or slightly longer of global cerebral ischemia (Hossmann et al., 1970) (Hossmann et al., 1974). I’m not aware of any studies finding that spontaneous global electrical activity is maintained after significantly longer than one hour of postmortem interval, in the absence of hypothermia or continued sustenance to the brain.

Local/cellular electrophysiology activity

Leonard 1991 studied hippocampal slices prepared from rat brains that were allowed to remain inside the skull after death for a delay ranging from 5 minutes to 3 hours (Leonard et al., 1991). They stimulated one area of the hippocampal slice and recorded from two areas. They found that the percentage of hippocampal slices that they were able to detect a electrophysiological response of > 1 mV decreased with increasing post-mortem interval. At a PMI of 5 or 30 minutes, around 65% of the slices had a detectable electrophysiological response, whereas this decreased after the 30 minute interval and was approximately 20% of the slices at 3 hours. The amplitude of the electrophysiological response also decreased with increasing PMI, with a half-life of approximately 1-2 hours. Overall, as the authors point out, meaningful neurobiological information can be obtained even if the brain tissue is not able to be collected immediately after death.

Charpak 1998 studied cortical, hippocampal, cerebellar, and thalamic brain slices prepared from rats whose brains were left in their skull for a variable amount of time after death (Charpak et al., 1998). The brains cooled naturally to room temperature of 20-24xC after around 25 minutes. The rats were killed by overdose on a ketamine/xylazine mixture, an overdose of pentobarbital, or decapitation. They did not detect any differences as a result of these different causes of death. After preparation, the brain slices were reoxygenated for 1 hour, which they reported was essential for tissue reactivation. Here are some of their results:

  • The first electrophysiology parameter that Charpak 1998 measured was the average resting membrane potential of layer V pyramidal cells. They found that the number they could record from decreased after 5-7 hours and they could not record from any cells after a postmortem delay of 12 hours. However, the average resting membrane potential remained constant for postmortem delays up to 9 hours, in the cells they were able to record from.

  • The second electrophysiology parameter that Charpak 1998 measured was the field potential resulting from stimulation of olfactory cortex brain slices. They found that the dendritic excitatory postsynaptic potential remained constant at PMI of up to 5 hours. At a PMI of 7 hours, they reported that barely any synaptic response was seen.

  • The third electrophysiology parameter that Charpak 1998 measured was intrinsic cell membrane and cell firing properties. They reported that these parameters were not altered across multiple brain regions for up to 5 hours of PMI.

  • Charpak 1998 also noted that in blind experiments, cortical slices were not able to be distinguished between a PMI of 0 hours and 5 hours. They describe the window of 5-6 as a “critical period” during which brain tissue can be reactivated, as long as the brain tissue is allowed to cool as it naturally does in the absence of blood flow.

Artemenko 2000 studied hippocampal slices prepared from rat brains that were kept inside of the skull a variable amount of time after death and kept either at 25xC or 30xC (Artemenko et al., 2000). They stimulated these slices along different neural pathways and measured the neuronal population responses. They found that the amplitude of the electrical response increased from 0 to 30 minutes and then began to decrease. When the brains were stored at 30xC, it took 2 hours for the slices to have an undetectable electrical response. When the brains were stored at 25xC, it took 3 hours. The decerease in amplitude of electrical response due to 90 minutes of PMI was dramatically decreased by the injection of neuroprotective chemicals prior to death.

Kramvis 2018 (Kramvis et al., 2018) studied human brain tissue donated after rapid autopsy with less than 10 hours PMI. In one donated brain with a PMI of 2 hours and 45 mins, they were able to record repetitive action potentials using patch clamp in pyramidcal cells in response to depolarizing steps of current. In two donated brains with post-mortem intervals between 3 and 4 hours, they were only able to elicit a single action potential in the majority of neurons studied. Beyond this, they were not able to identify any action potentials in the neurons that they could patch to record from. They were also able to identify synaptic currents in neurons recorded from brains with PMI less than 4 hours but not those with PMI greater than 4 hours.

Abbas 2020 studied field potentials in postmortem retinal tissue via electroretinography, which can be used to detect light responses of different populations of cells (Abbas et al., 2022). The retina has greater oxygen consumption per weight than any other tissue, so there is reason to think that it may be especially sensitive to anoxia. Here are some of their results:

  • Their first experiment measured retinal light responses in mice soon after death. They found that when the retina was left in its original location after death, the retinal light response was lost rapidly after circulatory death. They determined that the light response decreased in an exponential manner with increasing postmortem interval, with an exponential decay time constant of 1-2 minutes. (This is the amount of time it takes for the light response to decrease to approximately 36.8% of the original light response.)

  • Their second experiment was to waited a variable amount of time after death after the mice died, take the retinas out of the body, “revive” them by placing them into an oxygenated and nutrient rich medium, and only then measure the electrophysiological light response. They called this procedure “ex vivo”. They found that this procedure dramatically extended the amount of postmortem time it took for the light response to be lost, by greater than 10 fold. It took approximately 3 hours before the light responses became very small. The data still fit well to an exponential decay function (fig 3f).

  • One of their other experiments was to study the retinal light response in donated human eyes. They found that the retinal light response decreased with increasing PMI. At 5 hours PMI, they were still able to detect a retinal light response, albeit a weaker one. Notably, they found almost no bipolar cell light response in in the human samples, which are the second cells in the circuit of the retina. It makes sense that the bipolar cell response would degrade more quickly because it requires more components to be active at once: it requires a functional rod or cone, a functional bipolar cell, and a functional synapse between them. Whereas an initial rod or cone light response merely requires one cell to still be functional.

  • They also performed experiments to figure out why the retinal light response was degrading after a few hours of postmortem time. First they measured an apoptotic marker. They did not identify any significant amount of apoptosis in the retinal cells at 3 h PMI. This is not surprising because apoptosis generally does not seem to be a major contributor to cell death postmortem, in part because it is an active process and energy supplies have been largely depleted.

  • They then found that retinas incubated in low oxygen environments for 3 hours could not recover their light responses under restoration to normal conditions, whereas those incubated in low pH environments for 3 hours could. This suggests that, at least for retinas, it is the lack of oxygen more than the acid accumulation that most contributes to the postmortem loss of function that could not be reversed by incubation in restorative medium. (The authors call it “irreversible”, but that seems to me to be a stronger claim than their data suggest.)

Overall, most of the data suggests that electrophysiology responses lasts a few hours after death, from around 2 to 6 hours. As the PMI increases, there are diminished electrophysiological responses, suggesting that many of the neurons are no longer functional.

It may be illuminating to consider what types of structural components in cells are required for a detectable electrophysiological response. It requires ion channels and pumps to still be active and it requires cell membranes to not be so damaged that they lose their barrier function. It requires biomolecular distributions to still be broadly similar as in vivo. Anything beyond individual cell functions, such as excitatory postsynaptic potential stimulation, also requires that at least some portion of the synapses remain functional.

From an information perspective, an electrophysiology response is a relatively high bar. If even one part of the electrophysiology machinery is lost, then it is likely that the whole system may not have electrophysiological function. Whereas from a brain preservation perspective, the information in one part of the electrophysiology machinery is likely to be inferable from other structural information.

The idea that functional decline is not necessarily indicative of information loss is corroborated by the studies above suggesting that the postmortem interval window can be pushed further by reoxygenating and otherwise restoring neural tissue prior to the measurements. In other words, it may be possible to rescue electrophysiology phenotypes even with today’s technology, let alone the hypothetical technology of the future that would be required for revival from brain preservation.

One caveat here is that is that even if electrophysiology is present in postmortem brain tissue, it might not be the same as the electrophysiological responses seen in vivo. I haven’t seen any papers that try to correlate these two. That is certainly another complication when trying to interpret this type of data.

Breakdown of blood vessels

It’s not just neurons that are damaged by ischemia. Blood vessel cells are also rapidly damaged by the lack of oxygen and nutrients that occurs in ischemia. Circulatory system changes are often thought to be the major cause of the “5 minute rule” in neurology, whereby 5 minutes of lack of blood flow to the brain at room temperature is sufficient to cause the death of the organism (Leonard et al., 1991).

This is important because, in addition to neurovascular units being a potential source of cognition-relevant information storage, damage to blood vessels is also likely to hinder our attempts to perfuse preservative chemicals into blood vessels in brain preservation.

One example of blood vessel changes that can occur when the brain lacks blood flow is the no-reflow phenomenon. The idea behind the no-reflow phenomenon is that lack of blood flow causes primary damage to the blood vessels, which are then unable to sufficiently provide blood flow to the tissue even when it is restarted. This can then cause secondary damage to the rest of the brain tissue.

Ames 1968 is a classic study on the no-reflow phenomenon (Ames et al., 1968). In one set of experiments, they washed the blood out of rabbit brains and subjected them to 15 minutes of lack of blood flow. They found that these brains could still be perfused with carbon black, a substance with small particle size. But the circulatory system could no longer be perfused with blood, potentially because the red blood cells could not longer fit through the small arteries. Their findings suggested that one of the main factors limiting the recovery of blood flow after 5-15 minutes of ischemia was a reduction in the size of small blood vessels.

Ames 1968 also pointed out that the duration of ischemia that led to blood vessel damage corresponded very well to the duration of ischemia that leads to death of the animal. As a result, they suggested that these vascular changes may bring about the “point of no return” or be the first practically irreversible change.

A follow-up study by Fischer 1977 found that after around 30 minutes of lack of blood flow to the brain, around half of blood vessels showed extensive swelling of astrocytes that surround blood vessels (Fischer et al., 1977). Compared to blood vessels subjected to 3.5 minutes of ischemia, those subjected to 30 minutes of lack of blood flow also had significantly smaller diameters.

Another mechanism that has been suggested to help account for the no-reflow phenomenon is the contraction of pericytes, which are cells that wrap around blood vessel cells in capillaries (Heyba et al., 2019). It has been suggested that pericytes may undergo their own type of rigor mortis in the absence of blood flow that causes them to contract. This is an intriguing possibility because it fits with my conception of the importance of changes in the actin cytoskeleton after clinical death, but there is conflicting evidence about this and we need to avoid putting too much weight into alluring narratives.

Despite the breakdown in the brain’s circulatory system after clinical death, studies have shown that perfusion of the brain’s blood vessels is still possible to at least some extent (McFadden et al., 2019). But the PMI at which postmortem perfusion is still a useful procedure is poorly mapped. One investigator I spoke with who performed postmortem perfusion fixation suggested that doing this after 6 hours of PMI was not particularly helpful, as opposed to simple immersion fixation. Another suggested it was no longer worthwhile after 48 hours of PMI.

Cell viability

Cell death is a process in the same way that whole organism death is a process. There are plenty of quagmires related to the definition of cell death and when people refer to cell death they might be referring to a variety of different things.

Cell death is a functional process because it refers to the ability of the cell to function normally. Complicating this, cell death is often measured via structural proxies, such as measuring membrane integrity or DNA fragmentation.

There are many different types of cell death, but they generally fall into the categories of programmed cell death and non-programmed cell death.

Non-programmed cell death is generally known as necrotic cell death. This is somewhat misleading, because – according to a common definition of the term – cell death does not occur as result of necrosis (Martin et al., 1998). Instead, necrosis describes a particular end state of cellular morphology. Necrosis is a descriptive, not mechanistic term.

Non-programmed cell death can be thought of as a disintegrative or dissipative process due to an increase in the entropy of the cellular milieu and a lack of energy metabolism to curtail the increase in entropy (Lipton, 1999). The morphology that results reflects the mechanism of the injury (Martin et al., 1998).

In non-programmed cell death, the events that generally occur are that ion gradients are lost, few or no new macromolecules are synthesized, there is a loss of cell membrane activity, there is a massive influx of calcium and water, and eventually the cell structure degenerates. A common finding in studies of necrosis is that the overall shape of the cell tends to be maintained until the very end prior to cell dissolution (Martin et al., 1998).

Non-programmed cell death is the type of cell death that is most relevant following total lack of blood flow to the brain, because cells will not have enough energy available to carry out the active processes involved in programmed cell death. However, there may sometimes still be enough energy stores to perform some aspects of programmed cell death as well, so cell death after total lack of blood flow to the brain can potentially fall somewhere along an programmed versus non-programmed continuum.

One way to measure the amount of cell death in the brain is to quantify the number of live cells that can be cultured from the brain tissue. This is often called “viability”.

Viel 2001 studied rat brains that were left in the skull a variable amount of time after clinical death, stored either at room temperature or in ice (Viel et al., 2001). They then took brain samples and cultured the cells. Five days later, they determined the number of cells that retained both enzymatic activity and membrane integrity, which they defined as live cells.

When cells were cultured immediately after clinical death, approximately 30% of the cells remaining live at 5 days were neurons. The percentage of neurons among the live cells at 5 days steadily decreased with increasing post-mortem interval prior to culture. When the brains were stored at room temperature, this percentage was close to 0 at 8 hours of post-mortem interval. When the brains were stored at 4xC, this percentage was dramatically extended, to 4% at 32 hours of post-mortem interval. The percentage of total live cells also decreased with increasing post-mortem interval at both room temperature storage and 4xC storage.

The ability to culture cells from brain tissue after clinical death is an important topic in other areas of neuroscience research because it is useful to be able to culture brain cells to determine normal functioning and/or interrogate the pathophysiology of neurobiologic disease. However, in my opinion this type of viability study is not as important from a brain preservation perspective, for a few reasons.

First, calling a cell “dead” is problematic in the same way that calling an organism “dead” is problematic. Stages of cell death previously thought to be irreversible have been shown to be reversible, for example via restoring cells to optimal cell culture conditions (Tang et al., 2018). The whole point of brain preservation is that the technology to reverse the dying processes is likely to dramatically improve, so what we call cell “death” may no longer be such in the future.

Second, it is often the case that even in ideal circumstances, only a small percentage of cells are found to be “alive.” This seems to be clearly not enough for important aspects of cognition such as long-term memory recall, which likely relies on many or most of the cells in a particular ensemble to be functional. In other words, recovering isolated cells from the brain is not sufficient for cognition anyway, so it’s not a particularly useful metric of preservation quality.

Third, even if cells are found to be “alive” by various viability metrics, some other aspect of neural structure, such as their synaptic connectivity patterns, might still be lost. In this case, there would be loss of important cognitive information even if the cells were able to survive in culture.

In other words, loss of cell viability in the sense of “able to culture at least some of the brain cells” seems to be neither necessary nor sufficient for information theoretic death.

The other problem is that “viability” of cells across the brain is a very challenging goal and is unlikely to be reached for several decades, in my opinion. We can’t currently even get close to this. The state-of-the-art temperature possible for human organ preservation is typically via supercooling to high subzero temperatures. In the meantime, we can already preserve much of the structural information without viability.

In my opinion, much more important than whether a cell is “alive” or “dead”, from an information conservation perspective, is whether the likely structural components of engrams, such as cell membrane shape and biomolecules involved in ion flow, are preserved or inferable.

Breakdown of biomolecules during the dying process

It stands to reason that biomolecules that are attached to the cytoskeletal network directly through covalent bonds will retain their positions longer after death. One source notes that receptor proteins and other proteins embedded in the cell membrane, which is directly attached to the cytoskeleton, tend to be fairly stable even when measured some period of time after death (Waldvogel et al., 2006). On the other hand, biomolecules that are only indirectly constrained in place by the cyoskeletal mesh, such as neurotransmitter molecules and certain enzymes, are more likely to be lost during the dying process (Waldvogel et al., 2006).

Historically, postmortem biomolecule degradation has been thought to be enzyme-associated since the 1890s, when Salkowski used chloroform as bacteriostat and showed that various tissues still degraded (as cited in (Jensen, 1944), (Dernby, 1918). Salkowski described this process as “autodigestion”. In 1900, Jacoby did experiments to corroborate this phenomenon, describing non-bacterial postmortem degradation as “autolysis”, which is the term that stuck. As a result, the presence and activity of enzymes are also critical to understand during the amount of biomolecular degradation.

Nucleic acids

Here are some data points:

  • A large study finding no significant correlation between the RIN (an RNA quality metric) and the postmortem interval from 6-40 hours in human brains (Stan et al., 2006).
  • Another large study in human brains finding that the correlation between the PMI and RIN is weak (White et al., 2018):

A linear regression of data in Figure 1 showed that, while there was a significant inverse correlation between variables (p < 0.001), the value of PMI to predict RIN was very low (R2 = 0.013) in cortical tissue (Fig. 1A). A similar finding was observed in the cerebellum (Fig. 1B), where a linear regression showed that there was a trend (p = 0.06) for an inversion correlation between variables and the value of PMI to predict RIN was low (R2 = 0.003). Figure 1C shows that the effect of PMI on RIN was not evident, particularly before 36 hours.

  • RIN isn’t necessarily the best measure, though. For example, (Zhu et al., 2017) notes that:

these studies focused mainly on the correlation of RIN and PMI [post mortem interval]. However, RIN is largely derived based on the integrity of rRNA, and mRNA degradation may be distinctly different and gene specific, which is not completely reflected by the RIN.

Still, even when looking at mRNA specifically, (Zhu et al., 2017) (not looking at the brain) does not detect too many RNA transcripts that vary substantially in abundance by PMI.

  • This study finds that the brain has some of the slowest degradation of RNA postmortem compared to all other organs in the body (Bonadio et al., 2021):

Previous studies investigated the dynamics of gene expression after death in human liver samples, human prostate tissue, zebrafish and mice brains and human samples from 36 tissues recovered from the GTEx project, concluding that the brain (cortex and cerebellum) and spleen were the tissues presenting the higher RNA stability. This can be explained because the brain is protected by the neurocranium, thus being less exposed to exogenous RNases.

The brain may have relatively slower nucleic acid decomposition compared to other tissues because it is relatively immune-privileged and thus has relatively few endogenous nucleases (vs eg the GI tract, which needs to fight against RNA viruses).

  • This study reports good stability of mRNAs at 24 hours postmortem (Walker et al., 1992)

  • Notably, RNA can continue to be transcribed to some extent after organismal death, so the RNA present in cells after an extended PMI may not be the same as they were at the time of clinical death. This is known as the “thanatotranscriptome” (Pozhitkov et al., 2017). However, my impression from the literature is that this is a minor effect, and it is also a tricky thing to parse out methodologically.

  • One study found “Extensive postmortem stability of RNA from rat and human brain”, from 0 to 48 hours after death (Johnson et al., 1986).

  • In a cryonics context, Ralph Merkle has also summarized the literature showing the stability of mRNAs and proteins postmortem:

Morrison and Griffin said “We find that both rat and human cerebellar mRNAs are surprisingly stable under a variety of postmortem conditions and that biologically active, high-molecular-weight mRNAs can be isolated from postmortem tissue. … A comparison of RNA recoveries from fresh rat cerebella and cerebella exposed to different postmortem treatments showed that 83% of the total cytoplasmic RNAs present immediately postmortem was recovered when rat cerebella were left at room temperature for 16 h [hours] postmortem and that 90% was recovered when the cerebella were left at 4 degrees C for this length of time …. In neither case was RNA recovery decreased by storing the cerebella in liquid nitrogen prior to analysis. … Control studies on protein stability in postmortem rat cerebella show that the spectrum of abundant proteins is also unchanged after up to 16 h [hours] at room temperature….” Johnson et. al. in “Extensive Postmortem Stability of RNA From Rat and Human Brain” found that postmortem delays of as long as 48 hours “…failed to reveal degradation of the specific rat brain mRNAs during the postmortem period.” They also said “We find no effect of postmortem delay on RNA quality in both rat and human.”

  • In addition to RNA, DNA is also highly stable postmortem. One study shows it can be stable postmortem for months in the correct storage circumstances (Niemcunowicz-Janica et al., 2007).

  • A review of the gene expression literature by Robinson 2016 (Robinson et al., 2016) notes that people are often prejudiced about how many information is present in postmortem brains, before they look at the actual data:

The study of human post-mortem tissues is still essential for investigation of neurodegenerative disorders. It is a common perception that the long post-mortem delays (PMDs) that often limit the acquisition of brains after death are detrimental to brain quality, and therefore, prejudicial to the reliability of scientific data based on such materials. However, this viewpoint is largely anecdotal and has not been supported by scientific evidence.

Proteins

Here are some data points:

  • A 1921 study notes that brain proteins are peculiarily stable after death [Gibson et al. (1921)]. But that after a period of many days, the proteins do begin to break down into their amino acid forms.

  • A 1942 study notes that brain autolysis of proteins has long been thought to be slower than other organs (Sperry et al., 1942). However, this is assuming that the brain tissue is not acidotic.

Brain undergoes autolytic decomposition very slightly when measured by amino-acid production and when compared with most other tissues. The evidence of the reaction even under the most favorable conditions is so slight that it may easily be overlooked, or be incorrectly interpreted. It is because of this that the brain proteins are assumed to be peculiarly resistant to autolytic changes and not subject to those rather rapid fluctuations of mass which occur for example in the liver, and in other organs. The peculiar stability of brain and nerve proteins thus appears to relate itself to those functions of the brain which seem to demand a peculiarly stable physical structure, the functions of memory, habit, instinct, etc. While there can be no question about the stability of the protein framework of the brain cells, it does not necessarily follow that the autolytic mechanism is wanting, or is strikingly different from that of other organs. The outstanding fact of stability and permanence of brain cell structure may equally well be explained by the peculiarly perfect protection against local acidosis which the large blood supply of the brain insures.

The stability of brain structure depends not on any lack of proteolytic enzymes in the cells, but upon the fact that normally the brain proteins are not available substrata for the enzymes. They become available under the conditions that make liver proteins available for autolytic hydrolysis; namely, acidosis within the cells. That local acidoses do not frequently occur in brain tissue in spite of its large CO2 production, is due we believe to the exceptionally large blood supply to that organ, and to its ability to modify the respiratory and circulatorgi rate so as to prevent any accumulation of CO2 or other acid. By its extreme sensitiveness to increased H ion concentration and CO2 and by its position as master tissue of the body it automatically prevents just those conditions from arising within itself which would eventuate in its own autolytic disintegration… The autolytic disintegration of the delicate protein structures of brain tissue appears to be an irreversible phenomenon, and is accompanied by loss of such characteristic functions as memory, habit, motor control, and consciousness.

  • Another 1942 study notes that protein breakdown is not rapid or far reaching under most circumstances (Kies et al., 1942). However, brain proteinases appear to be very pH dependent:

For purely practical reasons, brain tissue is regarded by meat-packing technologists as prone to particularly rapid breakdown and spoilage. The presence of active autolytic agents in this tissue is thus to be expected. Yet previous investigators, while reporting the presence of a variety of hydrolytic enzymes, have not demonstrated the presence of any of them in remarkably large quantities. Nor has the autolysis of brain tissue, as judged by protein breakdown, appeared to be especially rapid or far reaching.

pH dependence of autolysis in brain and muscle; Source: (Kies et al., 1942)
pH dependence of autolysis in brain and muscle; Source: (Kies et al., 1942)
  • One study looked at the breakdown of calpain during global cerebral ischemia, finding that it underwent breakdown in the first 10 and 20 minutes after cardiac arrest (Neumar et al., 1996):

The average density of the 80‐kDa band decreased by 25 ± 4 (p = 0.008) and 28 ± 9% (p = 0.004) after 10 and 20 min of cardiac arrest, respectively, whereas the average density of the 78‐kDa band increased by 111 ± 50% (p = 0.02) after 20 min of cardiac arrest. No significant change in the density of the 76‐kDa band was detected. These results provide direct evidence for autolysis of brain μ‐calpain during cerebral ischemia

  • One study found that the measured levels levels of brain MBP decreased over 72 hours at both room temperature and 4°C, while levels of PLP and MAG did not (Barker et al., 2013).

  • This abstract reports that proteins can be successfully stained up to 7 days after death at room temperature in human brain tissue (Toupalík et al., 2001).

  • This well-controlled time series study of up to 24 hours finds that there is not much change in protein levels in the cortex (no significant changes) (MacDonald et al., 2019)

“We first investigated the effects of PMI on protein levels in human tissue within and between layers. In each analysis, the levels of only 3–10% of proteins were nominally correlated with PMI (p < 0.05) and none passed multiple hypothesis testing (q < 0.1) (Figure S6A–F). Second, similar to human tissue, in monkey tissue protein levels did not vary with PMI; levels of only 3% of proteins were nominally correlated with PMI and none passed multiple hypothesis testing (Figure S7). Finally, as a measure of the stability of protein localization, we examined the correlation between layer 5–3 differences in monkeys; only 9% were nominally correlated with PMI and none passed multiple hypothesis testing”

  • This study finds that the binding to NMDA receptors is stable for at least 72 hours postmortem (Geddes et al., 1986):

We examined the distribution and density of N-methyl-D-aspartate (NMDA) displaceable L-[3H]glutamate binding sites in human hippocampal samples obtained postmortem from Alzheimer’s disease (AD) patients and from age-matched controls. Binding to NMDA receptors was stable for at least 72 h postmortem, and the pharmacological profile corresponded to that described using electrophysiology… No change was observed in NMDA-displaceable L-[3H]gluta- mate binding up to 72 h postmortem (Fig. 1B). Thus, the distribution and density of N M D A receptors can be readily measured in the human brain obtained postmortem.

  • One study finds that Na+/K+-ATPase activity substantially decreases in the brain after death – compared to baseline, it was 65% decreased compared to baseline at 6 hours, 52% decreased at 24 h, and 84% decreased at 48 hours (da Fonseca et al., 2019).

  • Regarding protein post-translational modifications, many of these are thought to be more labile. For example, one source (Tashjian et al., 2019) notes that protein phosphorylation is considered a highly labile component of postmortem brain tissue:

However, while general immunohistochemical staining profiles for many proteins remain unchanged up to a PMI of 50 hours, distinctive western blot degradation patterns may be observed at that time, with protein phosphorylation states being the most labile. Furthermore, molecular profiling techniques have determined that, compared to DNA, prolonged PMIs may dramatically affect the integrity of phosphoproteins and RNA.

  • And one study finds empirically that there is a rapid effect of PMI on protein phosphorylation in postmortem brain tissue (Li et al., 2003).

  • My guess is that to the extent any labile protein phosphorylation states are critical for engram information, they are likely inferrable from other, less labile biomolecular information still present in the brain. However, this is an important area of uncertainty, and I would like to learn more about it.

  • One type of protein modification is the binding of an oligosaccharide to an asparagine residue of a protein, which is known as N-glycosylation. These bonds tend to be quite stable. For example, N-glycosylated protein modifications have been found intact in human bodies preserved for thousands of years through natural mummification (Ozcan et al., 2014).

Carbohydrates

Carbohydrate polymers can be classified into storage polysaccharides, which provide energy upon their breakdown, and structural polysaccharides, which form long-lasting components of cells and tissues molecules.

Glycogen is the primary storage polysaccharides in humans. It is degraded into glucose, a simple sugar. In amimals, glycosaminoglycans are an important group of structural polysaccharides. They include heparan sulfate, chondroitin sulfate, and hyaluronan. From a brain preservation perspective, what we tend to care about are the longer-lasting structural polysaccharides.

Changes in storage polysaccharides after death

The general finding from postmortem brain studies is that energetic metabolites are labile and are not preserved for a long period of time.

For example, almost all glucose is consumed within minutes after death. One study in mouse brains found that 15 seconds after clinical death, glucose levels were at 50% of baseline and 2 minutes after clinical death, glucose levels were at 11% of baseline [as cited in (Dienel, 2019). After this, glucose levels can actually increase, but this appears to be mostly an artifact due to the decomposition of polysaccharides that contain glucose as a subunit.

Glycogen levels have also been found to decrease rapidly during ischemia and to virtually disappear after 10-20 minutes of ischemia (Lowry et al., 1964). Consistent with this, another source reports that ischemia- or hypoxia-associated changes in tissues prior to fixatives entering cells are often responsible for glycogen losses in histology studies (Bullock, 1984).

The levels of other energy metabolites such as lactate and pyruvate are also dramatically altered after death. Postmortem lactate levels have been found to exceed the levels found in living brains by approximately 1000-fold (Dienel, 2019).

One study looks at the effect of postmortem delay on metabolites in macaque brain tissue collected at 30 mins compared to at 5-6 h after death (Kurochkin et al., 2019). Most metabolites were not affected, but 39/1405 metabolites were. The data is somewhat preliminary, however, as there are only 2 data points in the 5-6 hour group. In human brain tissue, they found that a postmortem interval between 3 and 50 hours explained 10% of the variance in metabolite levels, which was substantial, but less than other factors, including age, which explained 25% of the variation.

Overall, while there are exceptions, storage polysaccharides are rapidly consumed after clinical death. However, their levels tend to vary significantly in vivo as well, so they are likely not critical, non-redundant sites of information storage.

Changes in structural polysaccharides after clinical death

One study measured the immunoreactivity of one glycosaminoglycan, chondroitin sulfate, in human brain tissues with postmortem interval less than 24 hours (Yukawa et al., 2018). They found no association between the immunoreactivity and the PMI, suggesting that chondroitin sulfate molecules are not rapidly decomposed in the 24 hours following clinical death.

Other studies have measured structural polysaccharides in human brain tissue many hours after death, for example at postmortem intervals up to 24 hours (Huynh et al., 2019) or 32 hours (Raghunathan et al., 2020). That this type of study is possible suggests that glycosaminoglycans had not rapidly decomposed, but unfortunately these studies did not assess the quantitative association between PMI and polysaccharide levels.

Overall, there are not many studies so far evaluating the quantitative association between glycosaminoglycan levels and the postmortem interval. But at least at postmortem intervals of up to one day, the available evidence suggests that glycosaminoglycan levels do not rapidly decompose. This makes biophysical sense because glycosaminoglycans are generally crosslinked by structural proteins in the extracellular space of the brain into stable matrices (Richter et al., 2018).

Lipids

Lipids are defined as molecules that are insoluble in water but are soluble in organic solvents; that is, they are hydrophobic. There are many subclasses of lipids including fatty acids, phospholipids, sterols, and sphingolipids. They are a more complex, heterogeneous set of molecules than nucleic acids, proteins, and carbohydrates, because they are not as easily described in terms of their monomer building blocks. This heterogeneity makes it more difficult to summarize the changes in lipid levels following clinical death.

Here are some data points:

  • One study in whales found that a postmortem time of 15-22 hours did not substantially affect lipid contents (Tower et al., 1973):

Because the samples of whale brain were collected and frozen under conditions that were far from ideal, we deemed it essential to evaluate the potential complications of postmortem autolysis and inadequate preservation by freezing. Many hours intervened between the harpooning at sea and the sampling of the brains ashore. It is well known (JANSEN, 1952) that after death the body temperature of whales even in cold ocean waters decreases extremely slowly. Each of the brains which we obtained 15-22 h after harpooning was still distinctly warm when removed from the cranial cavity, so that extensive postmortem autolysis seemed inevitable. Hence we were surprised to find little evidence for postmortem autolysis. The gangliosides isolated from frozen and thawed cerebral cortical samples and separated by TLC consisted of GD;GT;GM in a normal pattern, with no detectable GM2 or GM3 (nomenclature of SVENNERHOLM, 1963) as would be expected if any extensive postmortem catabolism of native gangliosides had occurred. The absence of detectable amounts of haematoside (GM3) is noteworthy. Similarly the myelin basic proteins isolated from frozen and thawed cerebral cortical samples and subjected to electrophoresis on polyacrylamide gel at pH 2.4 migrated as a single band characteristic of myelin basic proteins from most mammalian species (MARTENSON 1969; 1971).

  • As with storage polysaccharides that can be broken down for the production of energy, the levels of lipids that can be broken down for energy such as free fatty acids tends to decrease rapidly after clinical death. For example, one study found that the levels of mouse brain free fatty acids, which can be catabolized for energy, increase rapidly after death (Aveldano et al., 1975). The breakdown appears to be an active process, perhaps brought about by the activation of phospholipase A by ischemia, because it can be inhibited by microwave irradiation of the brain.

  • This abstract suggests that there aren’t many changes in the composition of lipids for several days after death in sterile white matter (Lindlar et al., 1966):

Lipides change slightly during sterile autolysis of human white matter up to the 24 th day. From the 9th day on, there is a decrease of lipid-phosphorus caused by a breakdown of lecithin and plasmalogen accompanied by a comparable increase of free fatty acids (FFA) and aldehydes. In fresh white matter there are only traces of FFA. In the autolysing brain, contrary to several other autolysing organs, there are no lysophosphatides detectable by means of paper chromatography. Free and esterified cholesterol, cephalins, phosphoinositides, sphingomyelin, cerebrosides and sulfatides remain chromatographically constant during sterile autolysis. According to the above findings it can be concluded that postmortem autolysis up to several days duration has only neglectable effects on the lipid-composition of the white matter.

  • One study analyzed membrane lipid levels in the thalamus of frozen-thawed human brains with long postmortem intervals averaging 50 hours and ranging from 12 to 92 hours using thin-layer chromatography (Schmitt et al., 2004). They measured the levels of nine membrane lipids including the primary cell membrane phospholipid, phosphatidylcholine, and the primary myelin membrane component, sphingomyelin, and found that none of them were significantly associated with the postmortem interval:

Phosphatidylserine (PS) was increased (by 47.6%, p < .01) in the left thalamus of schizophrenic patients compared with control subjects, with the postmortem interval showing a significant influence (p < .03). Sphingomyelin (SM) was significantly decreased (by 47.5%, p < .01) in the patient group. This group also showed a marked decrease in both the total amount of GC1+2 (by 39.4%, p < .05) and phosphatidylcholine (PC) (by 27.8%, p < .05) (Figure 1, Figure 2); however, its breakdown product lysophosphatidylcholine (LPC) produced by phospholipase A2 (PLA2) activity did not differ between patients and control subjects. Phosphatidylethanolamine (PE), sulfatides (S), phosphatidylinositol (PI), and phosphatidylglycerol (PG) also showed no difference between control subjects and probands. PE, S, PC, LPC, SM, PS, PI, PG, and GC1+2 did not correlate with postmortem interval

  • Controlled studies in animal brains have shown that lipids begin to decompose after death, albeit relatively slowly. One study using radiolabeling in rat brains found a slight decrease in phosphatidylcholine to 92% of baseline levels at 30 minutes, while the levels of other complex lipids measured were not altered at that time point (Marion et al., 1979). This study also found a rapid rise in the levels of free fatty acids such as arachidonic acid, which increased 7x from baseline at 2 minutes and 21x from baseline at 15 minutes, as a result of the breakdown of a mixture of phospholipids.

  • One study found that after death there is a rapid increase in the amount of diacylglycerol in the brain (Lee et al., 1991). Diacylglycerol levels increased 5.5 times at 30 minutes, mostly due to the hydrolysis of phosphoinositides and to a lesser extent phosphatidylcholine.

Studies on postmortem brains (30 s to 30 min) showed a rapid increase in the total amount (from 40–50 nmol/g in 0 min to 210–290 nmol/g in 30 min) and in all the molecular species of DAG. Comparatively larger increases (seven‐ to 10‐fold) were found for the 18:0–20:4 and 16:0–20:4 species. Comparison of DAG species with the molecular species of different glycerolipids indicated that the rapid postmortem increase in content of DAG was mainly due to the breakdown of phosphoinositides. However, a slow but continuous breakdown of PC to DAG was also observed.

DAG has been shown to be present in small quantities in brain, and it has been shown that there is a rapid postmortem increase in its amount in brain (Banschbach and Geison, 1974; Aveldano and Bazan, 1975; Ikeda 1986). A preponderance of 18:O and 20:4 fatty acids in this DAG indicated that the rapid postmortem increase in its level was probably due to the hydrolysis of brain phosphoinositides

  • One study finds that most of the changes in lipid contents due to 10 minutes of global cerebral ischemia are delayed following reperfusion (Simão et al., 2013). The significant decreases in gangliosides, phospholipids, and cholesterol are not seen at 1 h after reperfusion, but only at 24 h or 7 d following reperfusion.

  • One study finds minimal changes in phospholipid levels up to 18 hours of PMI at a temperature of 5-10xC (Pearce et al., 2000):

Possible changes in PL composition with postmortem interval (PMI) in rat brain were examined. No significant changes were seen in PL headgroup or PC species composition with PMI at up to 18 hours A similar overall conclusion regarding the lack of effect of PMI can be drawn using the PLcm system, although the results are semiquantitative because of the partitioning of several PLs into the upper, aqueous layer. Comparison of the two methods by t‐test showed no composition differences for PC, SM, PEa, PEe, and CL/NAPE No new resonances appeared with increasing PMI. Simple statistical analysis of the results in Table 2 showed that, of the 13 PL measures, only dsPC appeared to correlate with PMI (r2 = 0.96, P = 0.02). However, when multiple correlations of dependent variables are performed, it is necessary to apply Bonferroni’s correction (essentially divide p by the number of correlations) when testing for significance (44). In the present case even the dsPC result falls out of statistical significance. We can safely say that the PL composition as measured by 31P NMR does not depend on PMI within the limits studied. This result will be quite useful in the application of these techniques to postmortem human brain samples, where a substantial PMI usually cannot be avoided

  • Lipids can be degraded at different rates in the PMI depending on the cause and circumstances present at the time of death. In one study, when barbituates were used, the PMI degradation was slower, possibly due to inhibition of phospholiapse (Jernerén et al., 2015 Jan-Feb):

A number of different euthanasia protocols are used in animal research. Yet, there is little discussion as to what extent the different methods may influence the interpretability of results and/or extrapolation between experiments. In the present study, the method of euthanasia markedly influenced the levels of FFAs in brain tissue. The group that was euthanized by ip injection of pentobarbital had decreased levels of all studied FFAs (15–44%) compared with the groups that were euthanized with decapitation or CO2 inhalation. This is in line with previous studies (Hattori et al., 1986, Shiu and Nemoto, 1981). Barbiturates such as pentobarbital are known to mediate their pharmacological effect by acting as agonists at the GABAA receptor but also alter the physiological properties of biological membranes (Harris & Schroeder, 1982). Furthermore, pentobarbital has been shown to inhibit phospholipase activity (Hattori, et al., 1987), likely via reduced protein kinase A activation (Dan’ura, Kurokawa, Yamashita, Yanagiuchi, & Ishibashi, 1986, Strokin et al., 2003). This probably explains the lower levels of FFAs in the group euthanized by injection of pentobarbital compared to the two other methods

Overall, it seems that there is a relatively fast active breakdown of lipids that occurs immediately following global cerebral ischemia, especially of energetic lipid species. However, the breakdown of structural lipids does not appear to be as rapid in human brain cases.

Small molecules

Many small molecules are rapidly altered in the postmortem interval. A few data points:

  • One study finds that creatinine levels in the brain increase within hours after death due to the cessation of metabolic activity (Gonzalez-Riano et al., 2017)

  • One study in retina tissue found that glutamate and taurine levels decreased in retinas 45 minutes after death (Abbas et al., 2022):

We quantified these differences in three cell types: photoreceptors (outer segments and somata), bipolar cells, and Müller cells. Taurine concentration is reduced in all three cell-subtypes after 45min postmortem. Both glutamine and glutamate are reduced in bipolar cells, photoreceptor cell somata, outer segments, and Müller glia at 45 min post-mortem (Extended Data Fig. 3). The loss of both the neurotransmitter required for transmission of photoreceptor potentials to the bipolar cells, as well as the loss of the key metabolite involved in the production of this neurotransmitter, could contribute to the rapid loss of photoreceptor-derived light response of ON bipolar cells after death.

  • One study finds that endogenous amines are lost within hours after death (Olson et al., 1973). Also reports some useful findings on how their histological visualization is lost during the PMI:

Our results regarding the cellular localization of endogenous monoamine stores post mortem are at variance with those of Constantinidis et al. and De La Torre claiming that more than 45 min after death ‘the endogenous amines are hopelessly diffused and histochemical visualization is no longer possible’, but the distribution of green fluorescence in the hypothalamus given by Constantinidis et al. is in agreement with our findings. It must be kept in mind, however, that although some areas were found to contain a relatively large number of nerve terminals post mortem in the present study, it may be assumed that these or other areas contained a larger amount of nerves in the living state. This is most probably the case especially in the cortical areas as indicated by the large increase in number of visible CA nerve terminals in these areas by the in vitro amine incubations. Smears of cortical areas can be obtained at neurosurgery and we know from such studies that the number of endogenously visible varicosities is considerable. However, also in slices obtained at neurosurgery there is an increase in the visible number of NA varicosities following a-methyl-NA incubation in vitro. It is clear that the most important post mortem change of the fluorescence histochemical picture is the gradual disappearance of endogenous intraneuronal amines with time. It was noted that nerve terminals close to the ventral surface of the brain and especially close to ependymal linings seemed to be best preserved.

Neurotransmitter levels

  • One study found no significant correlation between the PMI and levels of different neurotransmitter levels in human brain tissue: dopamine and its metabolites, amino acids including asparagine, aspartate, GABA, glutamine, and glutamate (Gerlach et al., 1996):

[F]ound no significant correlations between the post mortem interval and any of our neurotransmitter measurements in either the Parkinsonian or the control group. Further, it was shown previously in an animal model that simulates the decline in temperature under human autopsy conditions, that Glu and GABA are stable between 4 and 24 h post mortem.

  • One study showed good stability of amino acid neurotransmitters up to 24 hours postmortem in both an animal model and in human brains (Ellison et al., 1987):

The postmortem stability of amino acids was examined in an animal model simulating human autopsy conditions. Aspartate concentrations increased 15% between 4 and 24 h postmortem while γ-aminobutyric acid (GABA) concentrations rose 35% by 4 h but were stable thereafter. Glutamate and taurine were stable at all time points. The assay has been used to examine concentrations of neurotransmitter amino acids in 15 patients without neurological or psychiatric disease. Results agree well with previous work and knowledge of amino acid neurotransmitter pathways

Epigenetics

Many epigenetic markers seem to be generally preserved even after relatively long PMIs:

  • One study used pig and mouse models to study this, finding that DNA cytosine modifications and certain histone methylations maintained stability up to 72 hours post-mortem, while other histone methylations did so for at least 48 hours (Jarmasz et al., 2019). On the other hand, histone acetylations were less stable, with a noticeable decline within 24 hours post-mortem, particularly in large neuronal nuclei. Human brain samples had similar findings but seemingly slower kinetics, with histone acetylation changes evident at 4 to 5 days post-mortem.

  • One study on postmortem human brain found that most epigenetic signals are intact for at least up to 30 hours postmortem: (Huang et al., 2006):

We show that the bulk of nucleosomal DNA remains attached to histones during the first 30 h after death. Immunoprecipitation with antibodies against methylated histones was at least 10-fold more effective in unfixed, micrococcal nuclease-digested samples, in comparison to extracts prepared by fixation and sonication. Histone methylation differences across various genomic sites were maintained within a wide range of autolysis times and tissue pH. Therefore, immunoprecipitation of micrococcal nuclease-digested tissue extracts is a feasible approach to profile histone methylation at defined genomic loci in postmortem brain.

Cerebrospinal fluid

CSF is not a molecule, nor is it brain tissue, but it will be discussed here regardless. Generally speaking, CSF is relatively stable in the PMI, reportedly because it is in a protected environment. As long as the body is refrigerated after death, one study reported that cell count and cell morphology analysis was possible for up to 7 days postmortem (Bohnert et al., 2019). After this period, cells in the CSF tended to clot, overlap, have burst margins, or bacterial contamination:

It was found that standardized evaluation of cell count and cell morphology is possible under routine conditions up to 7 days post-mortem, if the body is promptly and sufficiently refrigerated after death and the cytospin preparation is made within 2 h after CSF collection. If the time interval between collection and preparation is longer, cellular clotting, bursting of cell margins, lack of differentiability, overlapping, and contamination such as reactive bacterial clusters and fungal spores could be observed. Consequently, the place of autopsy should be close to the laboratory ideally. Immediate cooling of the specimens to + 4 °C may briefly extend the time interval. Freezing is not recommended in order to avoid generating thermal artifacts in the cytospin preparation.

Variability factors

Pre-mortem standby

Much has been written about pre-mortem standby over the years. Here’s some thoughts by Mike Darwin on it, from 1995.

Overall, the remote stand-by model has proven extremely difficult to pull off successfully, while being extremely expensive. Done right, it will likely not be financially accessible to most people. This is not likely to change over time, because the primary factor is labor costs, meaning it is suspectible to Baumol’s cost disease. The only exception is if cryonics/brain preservation becomes more popular, in which the procedures are local and do not require remote stand-by.

Rapid autopsy

In some settings where there is strong collaboration between the research team and hospital staff, rapid autopsies can be performed. This can give a sense of how long it is possible for the brain to be preserved after death under ideal circumstances. For example, in one rapid autopsy protocol at Duke University reported in 1993, they had an average amount time from pronouncement to immersion of the brain tissue in cold fixative of 56 minutes, with a range of 22 to 90 minutes (Booze et al., 1993). The ability to perform a rapid autopsy is likely dependent on the circumstances of the death, which is a criterion for inclusion in the protocol. So it shouldn’t be considered a prospective data point of how quickly it would be practical to preserve the brain for an average case, but instead more of an ideal case.

Summary

How long of a postmortem interval is too long for information-theoretic death? Here are some loosely-held thoughts, all subject to revision:

References

Abbas, F., Becker, S., Jones, B. W., Mure, L. S., Panda, S., Hanneken, A. and Vinberg, F., Revival of Light Signalling in the Postmortem Mouse and Human Retina, Nature, vol. 606, no. 7913, pp. 351–57, June 2022. DOI: 10.1038/s41586-022-04709-x
Ames, A., Wright, R. L., Kowada, M., Thurston, J. M. and Majno, G., Cerebral Ischemia. II. The No-Reflow Phenomenon, The American Journal of Pathology, vol. 52, no. 2, pp. 437–53, February 1968.
Artemenko, D. P., Gerasimov, V. D. and Krishtal, O. A., Electrical Responses in Hippocampal Slices After Prolonged Global Ischemia: Effects of Neuroprotectors, Brain Research, vol. 863, no. 1–2, pp. 66–70, April 2000. DOI: 10.1016/s0006-8993(00)02097-7
Aveldano, M. I. and Bazan, N. G., Differential Lipid Deacylation During Brain Ischemia in a Homeotherm and a Poikilotherm. Content and Composition of Free Fatty Acids and Triacylglycerols, Brain Research, vol. 100, no. 1, pp. 99–110, accessed November 9, 2023, December 1975. DOI: 10.1016/0006-8993(75)90244-9
Bailey, D. M., Oxygen and Brain Death; Back from the Brink, Experimental Physiology, vol. 104, no. 12, pp. 1769–79, December 2019. DOI: 10.1113/EP088005
Barker, R., Wellington, D., Esiri, M. M. and Love, S., Assessing White Matter Ischemic Damage in Dementia Patients by Measurement of Myelin Proteins, Journal of Cerebral Blood Flow & Metabolism, vol. 33, no. 7, pp. 1050–57, accessed June 15, 2022, July 2013. DOI: 10.1038/jcbfm.2013.46
Bernstein, B. W., Chen, H., Boyle, J. A. and Bamburg, J. R., Formation of Actin-ADF/Cofilin Rods Transiently Retards Decline of Mitochondrial Potential and ATP in Stressed Neurons, American Journal of Physiology-Cell Physiology, vol. 291, no. 5, pp. C828–39, accessed November 9, 2023, November 2006. DOI: 10.1152/ajpcell.00066.2006
Bohnert, S., Ondruschka, B., Bohnert, M., Schuhmann, M. K. and Monoranu, C.-M., Post-Mortem Cerebrospinal Fluid Diagnostics: Cytology and Immunocytochemistry Method Suitable for Routine Use to Interpret Pathological Processes in the Central Nervous System, International Journal of Legal Medicine, vol. 133, no. 4, pp. 1141–46, July 2019. DOI: 10.1007/s00414-019-02050-z
Bonadio, R. S., Nunes, L. B., Moretti, P. N. S., Mazzeu, J. F., Cagnin, S., Pic-Taylor, A. and de Oliveira, S. F., Insights into How Environment Shapes Post-Mortem RNA Transcription in Mouse Brain, Scientific Reports, vol. 11, no. 1, p. 13008, accessed November 9, 2023, June 2021. DOI: 10.1038/s41598-021-92268-y
Booze, R. M., Mactutus, C. F., Gutman, C. R. and Davis, J. N., Frequency Analysis of Catecholamine Axonal Morphology in Human Brain. II. Alzheimer’s Disease and Hippocampal Sympathetic Ingrowth., Journal of the Neurological Sciences, 1993.
Boyer, D. M. and Harrington, A. R., Scaling of Bony Canals for Encephalic Vessels in Euarchontans: Implications for the Role of the Vertebral Artery and Brain Metabolism, Journal of Human Evolution, vol. 114, pp. 85–101, accessed November 9, 2023, January 2018. DOI: 10.1016/j.jhevol.2017.09.003
Bullock, G. R., The Current Status of Fixation for Electron Microscopy: A Review, Journal of Microscopy, vol. 133, no. 1, pp. 1–15, accessed April 4, 2019, 1984. DOI: 10.1111/j.1365-2818.1984.tb00458.x
Charpak, S. and Audinat, E., Cardiac Arrest in Rodents: Maximal Duration Compatible with a Recovery of Neuronal Activity, Proceedings of the National Academy of Sciences of the United States of America, vol. 95, no. 8, pp. 4748–53, accessed March 9, 2022, April 1998. DOI: 10.1073/pnas.95.8.4748
da Fonseca, C. A. R., Paltian, J., dos Reis, A. S., Bortolatto, C. F., Wilhelm, E. A. and Luchese, C., Na+/K+-ATPase, Acetylcholinesterase and Glutathione S-transferase Activities as New Markers of Postmortem Interval in Swiss Mice, Legal Medicine, vol. 36, pp. 67–72, accessed November 9, 2023, February 2019. DOI: 10.1016/j.legalmed.2018.11.003
Dernby, K. G., A STUDY ON AUTOLYSIS OF ANIMAL TISSUES, Journal of Biological Chemistry, vol. 35, no. 2, pp. 179–219, accessed November 9, 2023, August 1918. DOI: 10.1016/S0021-9258(18)86452-3
Dienel, G. A., Metabolomic Assays of Postmortem Brain Extracts: Pitfalls in Extrapolation of Concentrations of Glucose and Amino Acids to Metabolic Dysregulation In Vivo in Neurological Diseases, Neurochemical Research, vol. 44, no. 10, pp. 2239–60, accessed November 9, 2023, October 2019. DOI: 10.1007/s11064-018-2611-y
Dreier, J. P., Major, S., Foreman, B., Winkler, M. K. L., Kang, E.-J., Milakara, D., Lemale, C. L., et al., Terminal Spreading Depolarization and Electrical Silence in Death of Human Cerebral Cortex, Annals of Neurology, vol. 83, no. 2, pp. 295–310, accessed December 27, 2021, February 2018. DOI: 10.1002/ana.25147
Ellison, D. W., Beal, M. F. and Martin, J. B., Amino Acid Neurotransmitters in Postmortem Human Brain Analyzed by High Performance Liquid Chromatography with Electrochemical Detection, Journal of Neuroscience Methods, vol. 19, no. 4, pp. 305–15, accessed November 10, 2023, April 1987. DOI: 10.1016/0165-0270(87)90073-2
Etxeberria, F. and Serrulla, F., The Case of the Brains of La Pedraja: Forensic Sciences and Historical Memory in Spain, Mètode Science Studies Journal, vol. 10, pp. 109–17, accessed February 7, 2022, 2020.
Fischer, E. G., Ames, A., Hedley-Whyte, E. T. and O’Gorman, S., Reassessment of Cerebral Capillary Changes in Acute Global Ischemia and Their Relationship to the "No-Reflow Phenomenon", Stroke, vol. 8, no. 1, pp. 36–39, 1977. DOI: 10.1161/01.str.8.1.36
Font-Belmonte, E., González-Rodríguez, P. and Fernández-López, A., Necroptosis in Global Cerebral Ischemia: A Role for Endoplasmic Reticulum Stress, Neural Regeneration Research, vol. 15, no. 3, pp. 455–56, accessed November 9, 2023, September 2019. DOI: 10.4103/1673-5374.266054
Geddes, J. W., Chang-Chui, H., Cooper, S. M., Lott, I. T. and Cotman, C. W., Density and Distribution of NMDA Receptors in the Human Hippocampus in Alzheimer’s Disease, Brain Research, vol. 399, no. 1, pp. 156–61, December 1986. DOI: 10.1016/0006-8993(86)90611-6
Gerlach, M., Gsell, W., Kornhuber, J., Jellinger, K., Krieger, V., Pantucek, F., Vock, R. and Riederer, P., A Post Mortem Study on Neurochemical Markers of Dopaminergic, GABA-ergic and Glutamatergic Neurons in Basal Ganglia-Thalamocortical Circuits in Parkinson Syndrome, Brain Research, vol. 741, no. 1, pp. 142–52, accessed November 10, 2023, November 1996. DOI: 10.1016/S0006-8993(96)00915-8
Gibson, C. A., Umbreit, F. and Bradley, H. C., Studies of Autolysis: VII. Autolysis of Brain, Journal of Biological Chemistry, vol. 47, no. 2, pp. 333–39, accessed November 9, 2023, 1921.
Gonzalez-Riano, C., Tapia-González, S., García, A., Muñoz, A., DeFelipe, J. and Barbas, C., Metabolomics and Neuroanatomical Evaluation of Post-Mortem Changes in the Hippocampus, Brain Structure & Function, vol. 222, no. 6, pp. 2831–53, August 2017. DOI: 10.1007/s00429-017-1375-5
Hayman, J., Towards a More Accurate Estimation of the Time Since Death in Human Bodies Found Decomposed in Australian Conditions, accessed November 9, 2023, 2013. DOI: 10.25911/5d514dc542e08
Hayman, J. and Oxenham, M., Estimation of the Time Since Death in Decomposed Bodies Found in Australian Conditions, Australian Journal of Forensic Sciences, vol. 49, no. 1, pp. 31–44, accessed March 7, 2022, January 2017. DOI: 10.1080/00450618.2015.1128972
Heyba, M., Al-Abdullah, L., Henkel, A. W., Sayed, Z., Malatiali, S. A. and Redzic, Z. B., Viability and Contractility of Rat Brain Pericytes in Conditions That Mimic Stroke; an in Vitro Study, Frontiers in Neuroscience, vol. 13, accessed November 9, 2023, 2019. DOI: 10.3389/fnins.2019.01306
Hossmann, K. A. and Sato, K., Recovery of Neuronal Function After Prolonged Cerebral Ischemia, Science (New York, N.Y.), vol. 168, no. 3929, pp. 375–76, April 1970. DOI: 10.1126/science.168.3929.375
Hossmann, K. A. and Zimmermann, V., Resuscitation of the Monkey Brain After 1 h Complete Ischemia. I. Physiological and Morphological Observations, Brain Research, vol. 81, no. 1, pp. 59–74, November 1974. DOI: 10.1016/0006-8993(74)90478-8
Huang, H.-S., Matevossian, A., Jiang, Y. and Akbarian, S., Chromatin Immunoprecipitation in Postmortem Brain, Journal of Neuroscience Methods, vol. 156, no. 1–2, pp. 284–92, September 2006. DOI: 10.1016/j.jneumeth.2006.02.018
Huynh, M. B., Ouidja, M. O., Chantepie, S., Carpentier, G., Maïza, A., Zhang, G., Vilares, J., Raisman-Vozari, R. and Papy-Garcia, D., Glycosaminoglycans from Alzheimer’s Disease Hippocampus Have Altered Capacities to Bind and Regulate Growth Factors Activities and to Bind Tau, PLoS ONE, vol. 14, no. 1, p. e0209573, accessed November 9, 2023, January 2019. DOI: 10.1371/journal.pone.0209573
Ith, M., Bigler, P., Scheurer, E., Kreis, R., Hofmann, L., Dirnhofer, R., and Boesch, C., Identification of Metabolites Emerging During Autolysis and Bacterial Heterolysis of Decomposing Brain Tissue by 1H-MRS in Situ and in Vitro, Proceedings of the International Society for Magnetic Resonance in Medicine, vol. 10, p. 580, 2002.
Ith, Michael, Scheurer, E., Kreis, R., Thali, M., Dirnhofer, R. and Boesch, C., Estimation of the Postmortem Interval by Means of \(^{1}\)H MRS of Decomposing Brain Tissue: Influence of Ambient Temperature, NMR in Biomedicine, vol. 24, no. 7, pp. 791–98, August 2011. DOI: 10.1002/nbm.1623
Jarmasz, J. S., Stirton, H., Davie, J. R. and Del Bigio, M. R., DNA Methylation and Histone Post-Translational Modification Stability in Post-Mortem Brain Tissue, Clinical Epigenetics, vol. 11, no. 1, p. 5, accessed November 10, 2023, January 2019. DOI: 10.1186/s13148-018-0596-7
Jensen, L. B., MICROBIOLOGICAL PROBLEMS IN THE PRESERVATION OF MEATS, Bacteriological Reviews, vol. 8, no. 3, pp. 161–88, accessed November 9, 2023, September 1944. DOI: 10.1128/br.8.3.161-188.1944
Jernerén, F., Söderquist, M. and Karlsson, O., Post-Sampling Release of Free Fatty Acids - Effects of Heat Stabilization and Methods of Euthanasia, Journal of Pharmacological and Toxicological Methods, vol. 71, pp. 13–20, 2015 Jan-Feb. DOI: 10.1016/j.vascn.2014.11.001
Johnson, S. a., Morgan, D. g. and Finch, C. e., Extensive Postmortem Stability of RNA from Rat and Human Brain, Journal of Neuroscience Research, vol. 16, no. 1, pp. 267–80, accessed November 9, 2023, 1986. DOI: 10.1002/jnr.490160123
Kherani, Z. S. and Auer, R. N., Pharmacologic Analysis of the Mechanism of Dark Neuron Production in Cerebral Cortex, Acta Neuropathologica, vol. 116, no. 4, pp. 447–52, accessed August 29, 2022, October 2008. DOI: 10.1007/s00401-008-0386-y
Kies, M. W. and Schwimmer, S., OBSERVATIONS ON PROTEINASE IN BRAIN, Journal of Biological Chemistry, vol. 145, no. 2, pp. 685–91, accessed January 29, 2022, October 1942. DOI: 10.1016/S0021-9258(18)51311-9
Kramvis, I., Mansvelder, H. D. and Meredith, R. M., Chapter 22 - Neuronal Life After Death: Electrophysiologic Recordings from Neurons in Adult Human Brain Tissue Obtained Through Surgical Resection or Postmortem, in Handbook of Clinical Neurology, I. Huitinga and M. J. Webster, Eds., Elsevier, accessed November 9, 2023, pp. 319–33, 2018.
Krassner, M. M., Kauffman, J., Sowa, A., Cialowicz, K., Walsh, S., Farrell, K., Crary, J. F. and McKenzie, A. T., Postmortem Changes in Brain Cell Structure: A Review, Free Neuropathology, vol. 4, pp. 4–10, January 2023. DOI: 10.17879/freeneuropathology-2023-4790
Kurochkin, I., Khrameeva, E., Tkachev, A., Stepanova, V., Vanyushkina, A., Stekolshchikova, E., Li, Q., et al., Metabolome Signature of Autism in the Human Prefrontal Cortex, Communications Biology, vol. 2, no. 1, pp. 1–10, accessed November 9, 2023, June 2019. DOI: 10.1038/s42003-019-0485-4
Lee, C. and Hajra, A. K., Molecular Species of Diacylglycerols and Phosphoglycerides and the Postmortem Changes in the Molecular Species of Diacylglycerols in Rat Brains, Journal of Neurochemistry, vol. 56, no. 2, pp. 370–79, accessed November 9, 2023, 1991. DOI: 10.1111/j.1471-4159.1991.tb08161.x
Leonard, B. W., Barnes, C. A., Rao, G., Heissenbuttel, T. and McNaughton, B. L., The Influence of Postmortem Delay on Evoked Hippocampal Field Potentials in the in Vitro Slice Preparation, Experimental Neurology, vol. 113, no. 3, pp. 373–77, accessed November 9, 2023, September 1991. DOI: 10.1016/0014-4886(91)90028-B
Li, J., Gould, T. D., Yuan, P., Manji, H. K. and Chen, G., Post-Mortem Interval Effects on the Phosphorylation of Signaling Proteins, Neuropsychopharmacology, vol. 28, no. 6, pp. 1017–25, accessed November 9, 2023, June 2003. DOI: 10.1038/sj.npp.1300112
Lindblom, R. P. F., Tovedal, T., Norlin, B., Hillered, L., Englund, E. and Thelin, S., Mechanical Reperfusion Following Prolonged Global Cerebral Ischemia Attenuates Brain Injury, Journal of Cardiovascular Translational Research, vol. 14, no. 2, pp. 338–47, accessed February 23, 2022, April 2021. DOI: 10.1007/s12265-020-10058-9
Lindlar, F. and Güttler, R., Die Lipoide der weißen Hirnsubstanz während der Autolyse und bei der anämischen Erweichung, Acta Neuropathologica, vol. 6, no. 4, pp. 349–58, accessed November 9, 2023, December 1966. DOI: 10.1007/BF00688163
Lipton, P., Ischemic Cell Death in Brain Neurons, Physiological Reviews, vol. 79, no. 4, pp. 1431–1568, accessed January 29, 2022, January 1999. DOI: 10.1152/physrev.1999.79.4.1431
Lowry, O. H., Passonneau, J. V., Hasselberger, F. X. and Schulz, D. W., EFFECT OF ISCHEMIA ON KNOWN SUBSTRATES AND COFACTORS OF THE GLYCOLYTIC PATHWAY IN BRAIN, The Journal of Biological Chemistry, vol. 239, pp. 18–30, January 1964.
MacDonald, M. L., Favo, D., Garver, M., Sun, Z., Arion, D., Ding, Y., Yates, N., Sweet, R. A. and Lewis, D. A., Laser Capture Microdissectiontargeted Mass Spectrometry: A Method for Multiplexed Protein Quantification Within Individual Layers of the Cerebral Cortex, Neuropsychopharmacology, vol. 44, no. 4, pp. 743–48, accessed November 9, 2023, March 2019. DOI: 10.1038/s41386-018-0260-0
Madea, B., Importance of Supravitality in Forensic Medicine, Forensic Science International, vol. 69, no. 3, pp. 221–41, December 1994. DOI: 10.1016/0379-0738(94)90386-7
Marion, J. and Wolfe, L. S., Origin of the Arachidonic Acid Released Post-Mortem in Rat Forebrain, Biochimica Et Biophysica Acta (BBA) - Lipids and Lipid Metabolism, vol. 574, no. 1, pp. 25–32, accessed November 9, 2023, July 1979. DOI: 10.1016/0005-2760(79)90080-8
Martin, L. J., Al-Abdulla, N. A., Brambrink, A. M., Kirsch, J. R., Sieber, F. E. and Portera-Cailliau, C., Neurodegeneration in Excitotoxicity, Global Cerebral Ischemia, and Target Deprivation: A Perspective on the Contributions of Apoptosis and Necrosis, Brain Research Bulletin, vol. 46, no. 4, pp. 281–309, July 1998. DOI: 10.1016/s0361-9230(98)00024-0
McFadden, W. C., Walsh, H., Richter, F., Soudant, C., Bryce, C. H., Hof, P. R., Fowkes, M., Crary, J. F. and McKenzie, A. T., Perfusion Fixation in Brain Banking: A Systematic Review, Acta Neuropathologica Communications, vol. 7, no. 1, p. 146, September 2019. DOI: 10.1186/s40478-019-0799-y
McKeown, S. R., Postmortem Autolytic Response in Rat Brain Lysosomes, Journal of Neurochemistry, vol. 32, no. 2, pp. 391–96, accessed November 9, 2023, 1979. DOI: 10.1111/j.1471-4159.1979.tb00362.x
Miller, J. R. and Myers, R. E., Neurological Effects of Systemic Circulatory Arrests in the Monkey, Neurology, vol. 20, no. 7, pp. 715–24, July 1970. DOI: 10.1212/wnl.20.7.715
Miller, J. R. and Myers, R. E., Neuropathology of Systemic Circulatory Arrest in Adult Monkeys, Neurology, vol. 22, no. 9, pp. 888–904, September 1972. DOI: 10.1212/wnl.22.8.888
Moriya, F. and Hashimoto, Y., Postmortem Production of Ethanol and n-Propanol in the Brain of Drowned Persons, The American Journal of Forensic Medicine and Pathology, vol. 25, no. 2, pp. 131–33, June 2004. DOI: 10.1097/01.paf.0000127402.97576.42
Morton-Hayward, A. L., Thompson, T., Thomas-Oates, J. E., Buckley, S., Petzold, A., Ramsøe, A., O’Connor, S. and Collins, M. J., A Conscious Rethink: Why Is Brain Tissue Commonly Preserved in the Archaeological Record? Commentary on: Petrone P, Pucci P, Niola M, Et Al. Heat-induced Brain Vitrification from the Vesuvius Eruption in C.E. 79. N Engl J Med 2020;382:383-4. DOI: 10.1056/NEJMc1909867, STAR: Science & Technology of Archaeological Research, vol. 6, no. 1, pp. 87–95, accessed March 23, 2022, January 2020. DOI: 10.1080/20548923.2020.1815398
Mustata, M., Ritchie, K. and McNally, H. A., Neuronal Elasticity as Measured by Atomic Force Microscopy, Journal of Neuroscience Methods, vol. 186, no. 1, pp. 35–41, January 2010. DOI: 10.1016/j.jneumeth.2009.10.021
Neumar, R. W., Hagle, S. M., DeGracia, D. J., Krause, G. S. and White, B. C., Brain Mu-Calpain Autolysis During Global Cerebral Ischemia, Journal of Neurochemistry, vol. 66, no. 1, pp. 421–24, January 1996. DOI: 10.1046/j.1471-4159.1996.66010421.x
Niemcunowicz-Janica, A., Pepinski, W., Janica, J. R., Skawronska, M., J, J. J. and Koc-Zorawska, E., Typeability of AmpFlSTR SGM Plus Loci in Brain and Thyroid Gland Tissue Samples Incubated in Different Environments*, Journal of Forensic Sciences, vol. 52, no. 4, pp. 867–69, accessed November 9, 2023, 2007. DOI: 10.1111/j.1556-4029.2007.00465.x
Olson, L., Nystrom, B. and Seiger, A., Monoamine Fluorescence Histochemistry of Human Post Mortem Brain, Brain Research, vol. 63, pp. 231–47, accessed November 10, 2023, December 1973. DOI: 10.1016/0006-8993(73)90091-7
Ono, K., Kim, S. O. and Han, J., Susceptibility of Lysosomes to Rupture Is a Determinant for Plasma Membrane Disruption in Tumor Necrosis Factor Alpha-Induced Cell Death, Molecular and Cellular Biology, vol. 23, no. 2, pp. 665–76, accessed November 9, 2023, January 2003. DOI: 10.1128/MCB.23.2.665-676.2003
Ozcan, S., Kim, B. J., Ro, G., Kim, J.-H., Bereuter, T. L., Reiter, C., Dimapasoc, L., et al., Glycosylated Proteins Preserved over Millennia: N-glycan Analysis of Tyrolean Iceman, Scythian Princess and Warrior, Scientific Reports, vol. 4, no. 1, p. 4963, accessed November 9, 2023, May 2014. DOI: 10.1038/srep04963
Pearce, J. M. and Komoroski, R. A., Analysis of Phospholipid Molecular Species in Brain by (31)P NMR Spectroscopy, Magnetic Resonance in Medicine, vol. 44, no. 2, pp. 215–23, August 2000. DOI: 10.1002/1522-2594(200008)44:2<215::aid-mrm8>3.0.co;2-n
Pozhitkov, A. E., Neme, R., Domazet-Lošo, T., Leroux, B. G., Soni, S., Tautz, D. and Noble, P. A., Tracing the Dynamics of Gene Transcripts After Organismal Death, Open Biology, vol. 7, no. 1, p. 160267, January 2017. DOI: 10.1098/rsob.160267
Pulsinelli, W. A., Selective Neuronal Vulnerability: Morphological and Molecular Characteristics, Progress in Brain Research, vol. 63, pp. 29–37, 1985. DOI: 10.1016/S0079-6123(08)61973-1
Raghunathan, R., Hogan, J. D., Labadorf, A., Myers, R. H. and Zaia, J., A Glycomics and Proteomics Study of Aging and Parkinson’s Disease in Human Brain, Scientific Reports, vol. 10, no. 1, p. 12804, accessed November 9, 2023, July 2020. DOI: 10.1038/s41598-020-69480-3
Ravindran, S. and Kurian, G. A., Eventual Analysis of Global Cerebral Ischemia-Reperfusion Injury in Rat Brain: A Paradigm of a Shift in Stress and Its Influence on Cognitive Functions, Cell Stress & Chaperones, vol. 24, no. 3, pp. 581–94, accessed November 9, 2023, May 2019. DOI: 10.1007/s12192-019-00990-4
Richter, R. P., Baranova, N. S., Day, A. J. and Kwok, J. C., Glycosaminoglycans in Extracellular Matrix Organisation: Are Concepts from Soft Matter Physics Key to Understanding the Formation of Perineuronal Nets?, Current Opinion in Structural Biology, vol. 50, pp. 65–74, accessed November 9, 2023, June 2018. DOI: 10.1016/j.sbi.2017.12.002
Robinson, A. C., Palmer, L., Love, S., Hamard, M., Esiri, M., Ansorge, O., Lett, D., et al., Extended Post-Mortem Delay Times Should Not Be Viewed as a Deterrent to the Scientific Investigation of Human Brain Tissue: A Study from the Brains for Dementia Research Network Neuropathology Study Group, UK, Acta Neuropathologica, vol. 132, no. 5, pp. 753–55, accessed November 9, 2023, November 2016. DOI: 10.1007/s00401-016-1617-2
Routtenberg, A. and Tarrant, S., Synaptic Morphology and Cytoplasmic Densities: Rapid Post-Mortem Effects, Tissue & Cell, vol. 6, no. 4, pp. 777–88, 1974. DOI: 10.1016/0040-8166(74)90015-9
Schmitt, A., Wilczek, K., Blennow, K., Maras, A., Jatzko, A., Petroianu, G., Braus, D. F. and Gattaz, W. F., Altered Thalamic Membrane Phospholipids in Schizophrenia: A Postmortem Study, Biological Psychiatry, vol. 56, no. 1, pp. 41–45, July 2004. DOI: 10.1016/j.biopsych.2004.03.019
Shrestha, R., Kanchan, T. and Krishan, K., Methods of Estimation of Time Since Death, in StatPearls, Treasure Island (FL): StatPearls Publishing, accessed November 9, 2023, 2023.
Siesjö, B. K., Cell Damage in the Brain: A Speculative Synthesis, Journal of Cerebral Blood Flow and Metabolism: Official Journal of the International Society of Cerebral Blood Flow and Metabolism, vol. 1, no. 2, pp. 155–85, 1981. DOI: 10.1038/jcbfm.1981.18
Simão, F., Matté, A., Breier, A. C., Kreutz, F., Trindade, V. M. T., Netto, C. A. and Salbego, C. G., Resveratrol Prevents Global Cerebral Ischemia-Induced Decrease in Lipid Content, Neurological Research, vol. 35, no. 1, pp. 59–64, accessed November 9, 2023, January 2013. DOI: 10.1179/1743132812Y.0000000116
Sperry, W. M., Brand, F. C. and Copenhaver, W. M., THE BEHAVIOR OF LIPIDS DURING AUTOLYSIS OF LIVER AND BRAIN, Journal of Biological Chemistry, vol. 144, no. 2, pp. 297–306, accessed November 9, 2023, July 1942. DOI: 10.1016/S0021-9258(18)72508-8
Stan, A. D., Ghose, S., Gao, X.-M., Roberts, R. C., Lewis-Amezcua, K., Hatanpaa, K. J. and Tamminga, C. A., Human Postmortem Tissue: What Quality Markers Matter?, Brain Research, vol. 1123, no. 1, pp. 1–11, accessed January 29, 2022, December 2006. DOI: 10.1016/j.brainres.2006.09.025
Tabet, A., Mommer, S., Vigil, J. A., Hallou, C., Bulstrode, H. and Scherman, O. A., Mechanical Characterization of Human Brain Tissue and Soft Dynamic Gels Exhibiting Electromechanical Neuro-Mimicry, Advanced Healthcare Materials, vol. 8, no. 10, p. e1900068, May 2019. DOI: 10.1002/adhm.201900068
Tang, H. M. and Tang, H. L., Anastasis: Recovery from the Brink of Cell Death, Royal Society Open Science, vol. 5, no. 9, p. 180442, accessed November 9, 2023, September 2018. DOI: 10.1098/rsos.180442
Tashjian, R. S., Williams, R. R., Vinters, H. V. and Yong, W. H., Autopsy Biobanking: Biospecimen Procurement, Integrity, Storage, and Utilization, Methods in Molecular Biology (Clifton, N.J.), vol. 1897, pp. 77–87, accessed January 29, 2022, 2019. DOI: 10.1007/978-1-4939-8935-5_8
Teo, C. H., Hamzah, N. H., Hing, H. L. and Hamzah, S. P. A. A., Decomposition Process and Post Mortem Changes: Review, Sains Malaysiana, vol. 43, no. 12, p. 1873, accessed November 9, 2023, 2014.
Toupalík, P., Bouska, I. and Jezková, J., [Effect of autolysis on histochemical examinations of the central nervous system], Soudni lekarstvi, vol. 46, no. 2, pp. 18–20, April 2001.
Tower, D. B. and Young, O. M., Interspecies Correlations of Cerebral Cortical Oxygen Consumption, Acetylcholinesterase Activity and Chloride Content: Studies on the Brains of the Fin Whale (Baaenoptera Physalus) and the Sperm Whale (Physeter Catodon), Journal of Neurochemistry, vol. 20, no. 2, pp. 253–67, accessed November 9, 2023, 1973. DOI: 10.1111/j.1471-4159.1973.tb12125.x
Viel, J. J., McManus, D. Q., Cady, C., Evans, M. S. and Brewer, G. J., Temperature and Time Interval for Culture of Postmortem Neurons from Adult Rat Cortex, Journal of Neuroscience Research, vol. 64, no. 4, pp. 311–21, May 2001. DOI: 10.1002/jnr.1081
Vrselja, Z., Daniele, S. G., Silbereis, J., Talpo, F., Morozov, Y. M., Sousa, A. M. M., Tanaka, B. S., et al., Restoration of Brain Circulation and Cellular Functions Hours Post-Mortem, Nature, vol. 568, no. 7752, pp. 336–43, April 2019. DOI: 10.1038/s41586-019-1099-1
Waldvogel, H. J., Curtis, M. A., Baer, K., Rees, M. I. and Faull, R. L. M., Immunohistochemical Staining of Post-Mortem Adult Human Brain Sections, Nature Protocols, vol. 1, no. 6, pp. 2719–32, 2006. DOI: 10.1038/nprot.2006.354
Walker, E. and McNicol, A. M., In Situ Hybridization Demonstrates the Stability of mRNA in Post-Mortem Rat Tissues, The Journal of Pathology, vol. 168, no. 1, pp. 67–73, September 1992. DOI: 10.1002/path.1711680112
Weickenmeier, J., Kurt, M., Ozkaya, E., de Rooij, R., Ovaert, T. C., Ehman, R. L., Butts Pauly, K. and Kuhl, E., Brain Stiffens Post Mortem, Journal of the Mechanical Behavior of Biomedical Materials, vol. 84, pp. 88–98, August 2018. DOI: 10.1016/j.jmbbm.2018.04.009
White, K., Yang, P., Li, L., Farshori, A., Medina, A. E. and Zielke, H. R., Effect of Postmortem Interval and Years in Storage on RNA Quality of Tissue at a Repository of the NIH NeuroBioBank, Biopreservation and Biobanking, vol. 16, no. 2, pp. 148–57, accessed January 29, 2022, April 2018. DOI: 10.1089/bio.2017.0099
Yukawa, T., Iwakura, Y., Takei, N., Saito, M., Watanabe, Y., Toyooka, K., Igarashi, M., et al., Pathological Alterations of Chondroitin Sulfate Moiety in Postmortem Hippocampus of Patients with Schizophrenia, Psychiatry Research, vol. 270, pp. 940–46, December 2018. DOI: 10.1016/j.psychres.2018.10.062
Zhu, Y., Wang, L., Yin, Y. and Yang, E., Systematic Analysis of Gene Expression Patterns Associated with Postmortem Interval in Human Tissues, Scientific Reports, vol. 7, no. 1, p. 5435, accessed January 9, 2022, July 2017. DOI: 10.1038/s41598-017-05882-0