GCs and the Death of Hippocampal Neurons

GC Endangerment. A body of work initiated by Robert Sapolsky and colleagues in the 1980s demonstrated the ability of transient bursts of GC excess to endanger hippocampal neurons, compromising their ability to survive a coincident neurological insult. As such, the greater the GC levels at the time of an insult, the greater the extent of hippocampal damage.

Such insults include epileptic seizures, hypoxia-ischemia, hypoglycemia and other metabolic insults, oxidative insults, the ß-amyloid peptide (the fragment of the amyloid precursor protein that has been implicated in the neuron death of Alzheimer’s disease), and gp120 (the coat protein of the AIDS virus that is thought to play a role in the neurodegeneration seen in AIDS-related dementia).

This endangerment is of a considerable magnitude, with stress levels of GCs causing manyfold increases in insult-induced damage in some cases.

Features of this endangerment include the following:
(1) it is mediated by GRs, rather than MRs;
(2) while most dramatic in the hippocampus, the endangerment can also occur in the cortex and striatum;
(3) it is a direct GC effect in the hippocampus (rather than secondary to some peripheral GC action), as the same phenomenon occurs in hippocampal cultures and tissue slices; and
(4) it consists of GCs increasing the amount of insult-induced necrotic death, rather than an increase in the amount of apoptotic death (a form of programmed cell death).

In vitro models have made it possible to uncover some of the mechanisms underlying the endangerment. One realm concerns the disruptive effects of GCs upon neuronal energetics. GCs inhibit glucose transport in many peripheral tissues (as part of the strategy to divert energy to exercising muscle during a stressful crisis); this effect is due to a reduction in the number of glucose transporter molecules in the cell membrane. GCs cause a similar inhibition in the hippocampus, leaving neurons energetically vulnerable, thereby worsening insult-induced declines in energy substrates (such as ATP).

Such insults cause neurons to release damaging amounts of the excitatory neurotransmitter glutamate. This, in turn, causes a wave of calcium influx in the postsynaptic neuron, which causes an array of calcium-dependent degenerative events; of most significance is the generation of oxygen radicals. It is quite costly for a neuron to contain the excesses of glutamate, calcium, and oxygen radicals, and the mild energetic vulnerability caused by GCs makes each of these steps spiral a bit more out of control, increasing the likelihood of the neuron succumbing.

In addition to these energetic features, GC endangerment probably also arises from the demonstrated abilities of the hormone to increase calcium currents in hippocampal neurons, to decrease the activity of calcium-extrusion mechanisms, to suppress the levels or efficacy of neurotrophins (neuronal growth factors), and to disrupt some of the defenses normally mobilized by neurons during neurological insults.

At present, these data are derived exclusively from rodent and tissue culture studies. Should subsequent work demonstrate GC endangerment in the primate hippocampus, they have a number of disturbing implications. Numerous individuals are administered high-dose GCs to control autoimmune or inflammatory disorders, and this might well exacerbate the damaging effects of some neurological crisis. Plus, GCs are often administered after strokes in order to decrease brain edema (swelling).

Most studies show that GCs are less effective at reducing such edema than nonsteroid anti-inflammatory compounds; the present studies suggest that the use of GCs might even worsen the neurological outcome. Finally, stroke, seizure, and cardiac arrest all cause massive GC secretion; such endogenous GC release might add to damage. In effect, what we have learned to view as the typical extent of neurological damage caused by these insults might well reflect the damaging effects of the maladaptive hypersecretion of GCs at that time.

GC Neurotoxicity. Independent of a coincident neurological insult, GCs can directly influence the life and death of a hippocampal neuron. In an unexpected finding, Robert Sloviter and colleagues reported that GCs are required for the survival of dentate gyrus neurons. In the complete absence of the hormone, such neurons undergo apoptosis within days, and very low levels of GCs, via MR occupancy, prevent this effect.

A separate literature showed that truly prolonged exposure to elevated GC levels could directly kill hippocampal neurons. Studies initiated by Phil Landfield and colleagues showed that the loss was centered in the CA3 region of the hippocampus and was highly relevant to hippocampal aging. That region of the hippocampus loses neurons with age, and these studies suggested that it is the cumulative extent of GC exposure over the lifetime that acts as a major determinant of the extent of neuron loss (as well as of hippocampal-dependent memory problems) in old age.

As such, stress has the capacity to accelerate hippocampal aging, whereas behavioral or social manipulations that reduce cumulative GC exposure can delay such aging. The finding of GC- and of stress-induced neurotoxicity was replicated by other investigators, including in the nonhuman primate hippocampus.

The mechanisms underlying such neurotoxicity are not well understood. The assumption by most in the field, however, is that it is on a continuum with GC neuroendangerment - in effect, if a day of GC overexposure can compromise the ability of a neuron to survive a massive neurological insult, months of overexposure can compromise its ability to survive the minor metabolic challenges of everyday life, especially during aging. The difficult studies required to test this idea remain to be carried out.

Despite the replication of the finding of GC neurotoxicity, the phenomenon is more controversial than the other adverse GC effects that have been discussed. The disagreements have focused on a number of issues. The basic finding of neuron loss during hippocampal aging has been called into question by some researchers counting neurons with the technique of stereology.

Some investigators suggest that overt neurotoxicity can be caused only by massive amounts of ever-shifting stressors (to avoid habituation of the system) or by pharmacological GC exposure, questioning the relevance of the phenomenon to normative aging. GC hypersecretion and hippocampal neuron loss in old age are quite variable among rodents. There appear to be strains in which neither occurs to any dramatic extent.

Moreover, dramatic differences in the extent of successful aging can occur within strains. Michael Meaney and colleagues, for example, have shown that remarkably subtle differences in the quality of early maternal care can change the entire trajectory of hippocampal and adrenocortical aging in a rat, in some cases completely avoiding the neuron loss or GC hypersecretion; thus, neither component of that cascade model is an inevitable part of aging. Finally, to the extent that primates and humans hypersecrete GCs in old age, it is a nonlinear phenomenon, emerging only in extreme old age or when coincident with an insult (such as major depression or Alzheimer’s disease).