Farid Hamzei-Sichani

I study neuronal gap junctions using thin-section transmission electron microscopy (TEM) and freeze-fracture replica immunogold labeling (FRIL) electron microscopy (through a collaboration with Prof. Patrick Hof at Mount Sinai School of Medicine and Prof. John Rash at Colorado State University). The primary focus of my studies is the hippocampal formation, mainly the hippocampal mossy fiber axons and the mossy fiber-CA3 synapses.

Gap junctions between principal cell axons have been predicted to exist, based on modeling, electrophysiological, and pharmacological data, and recognized to be essential for the generation of fast (>20 Hz) oscillations.

In a more general sense, the existence of gap junctions between principal cell axons is entirely counter-intuitive. One imagines that "cross-talk" between principal cell axons would degrade the precision of information processing in single neurons, and would predispose to epilepsy.

Figure 1: Electron micrograph of the stratum lucidum in an adult rat hippocampus showing a bundle of mossy fiber axons (mfa) cut in various angles. Dendrties of CA3 pyramidal neurons (Dend) are surrounded by mossy fiber axons or their terminals (mfb). Scale bar is 500 nm.
Figure 2: Electron micrograph of a giant mossy fiber terminal in the stratum lucidum of an adult rat hippocampus. The mossy fiber terminal (mfb) contains clear and dense-core vesicles and is perforated by at least two dendritic spines (asterisks). Arrowheads mark the asymmetric glutamatergic synapses onto branched spines of CA3 pyramidal neurons. Note the presence of mitochondria in the mossy fiber terminal and the spine apparatus within the dendritic spines. Scale bar is 100 nm.

Nevertheless, a preponderance of evidence points to the existence of axoaxonic gap junctions under physiological and pathological conditions, providing pathways for the axons to communicate with each other (and produce oscillations) independent of soma and dendritic compartments. The existence of such gap junctions requires the Neuroscience community to re-evaluate/redefine one of its most basic assumptions, i.e. the Neuron Doctrine.

Historical Perspective

Prof. Roger Traub (my Ph.D. advisor at SUNY Downstate Medical Center) proposed the hypothesis of gap junctional coupling between axons of cortical principal neurons based on the shape of putative coupling potentials (spikelets or fast prepotentials) during low-calcium ~200 Hz ripples in vitro (Draguhn et al., Nature, 1998). Subsequently, Schmitz and colleagues (Neuron, 2001) provided electrophysiological and dye-coupling evidence for axoaxonic coupling in CA1 pyramidal neurons as well as in dentate granule cells. It was shown on theoretical and modeling grounds that axonal coupling could indeed account for very fast (>70 Hz) network oscillations (Traub et al., Neuroscience, 1999); and that axonal very fast oscillations were critical in the generation of persistent gamma oscillations (Traub et al., PNAS, 2003), an in vitro experimental phenomenon first described in Nature (Fisahn et al., 1998). Very fast oscillations (~80-100 Hz) have been observed in the dentate gyrus under conditions when chemical synapses are blocked (Towers et al., J. Neurophysiol., 2002), providing physiological evidence that axoaxonic coupling might be expected to occur in this structure.

Clinical Perspective

Electrical coupling between axons could have profound consequences for clinical epilepsy (Traub et al., Epilepsia, 2001). Epileptogenesis in the dentate gyrus, as studied after kainate-induced neurotoxicity, has been an interesting and important model system for the study of lesion-induced seizure foci; the emphasis to date has been on the sprouting of mossy fibers, but gap junctions could clearly play a role as well.