GCs and the Function and Structure of Hippocampal Neurons

As just discussed, mild stressors, via MR occupancy, enhance hippocampal-dependent cognition and LTP, whereas major stressors, via GR occupancy, disrupt both. The mechanisms underlying these actions are well understood.

Major stressors and prolonged GC exposure can also cause structural changes in hippocampal neurons. Neurons are highly arborized cells, with information flowing in from neighboring neurons via a dense tree of processes called dendrites, and with information flowing out through a major cable called the axon. The extent of dendritic arborization - the length and complexity of processes - is now recognized to be highly dynamic, being remodeled by a variety of local, systemic, and even experiential influences.

Work initiated by McEwen and colleagues has shown that in the rat, sustained exposure to stress or to excessive GC levels over the course of weeks can cause atrophy of dendrites in hippocampal neurons. Reflecting the dynamic nature of dendritic remodeling, the atrophy appears to gradually reverse when the stress or GC overexposure stops. Subsequent work from this group and others has shown that the same atrophy occurs in the nonhuman primate hippocampus as well.

It is the GR receptor that appears to mediate the deleterious GC effects on the dendrites. Moreover, GCs appear to cause the atrophy through the glutamate neurotransmitter system. As evidence, atrophy can be prevented by blocking NMDA glutamate receptors in the hippocampus or by the use of drugs that blunt glutamate release (such as the seizure suppresser phenytoin). Thus, the phenomena of GC- induced dendritic atrophy and of neuroendangerment appear to be linked mechanistically.

What are the consequences of this dendritic atrophy? There are at least three possibilities. First, part of the staggering complexity of the nervous system comes from the multiplicity of synaptic connections between neurons, and a typical neuron can form thousands of synapses with neighbors both near and far. The existence of such neuronal networks is obviously dependent upon an intact dendritic tree, and dendritic atrophy is likely to disrupt and simplify networks within the hippocampus.

Stress or GC overexposure on the scale needed to cause dendritic atrophy will also disrupt cognition, and it seems plausible that the two are interrelated. Second, as will be reviewed shortly, GCs and stress can make hippocampal neurons less likely to survive coincident neurological insults. It has been suggested that the dendritic atrophy represents, in effect, a hibernation, of a neuron, an involution that protects it from a coincident insult.

Third, in contrast, others have suggested that the atrophy represents a marker of the neuron being made more vulnerable, rather than less, to an insult. While it remains unclear whether the first two models are correct, most existing evidence supports the last.