Introduction

Nerve and muscle cells that partially lose their synaptic inputs experience one of the following: (1) postsynaptic transcellular degeneration, if the differentiation is severe; (2) cellular survival, but with fewer synaptic inputs than before; or (3) selected denervated sites becoming renervated from “new” axon connections (185). The process whereby axons and corresponding synapses are replaced after partial differentiation is termed reactive synaptogenesis (35).

The concept that after partial synaptic loss, nearby undamaged axons sprout new branches to reform synaptic contacts has existed for years. It originated in studies of the peripheral nervous system. Haighton, in 1795, was perhaps the earliest to suggest that after denervation, surviving peripheral motor nerves could reform functional connections and might even hyperinnervate organs such as the stomach and vocal cords, thus explaining postinjury dysfunctional physical signs (68). After incomplete proximal motor nerve transection, Exner (51) in 1885 observed that muscular contraction appeared prior to nerve fiber regeneration, and concluded that this was from “new” local collateral nerve growth. Likewise, Kennedy (85) in 1897 cautioned that one explanation for studies purporting to show peripheral nerve regeneration might be that local undamaged nerves sprouted into denervated sites. Subsequently, studies by Lugaro (104), Ramón y Cajal (161), Van Harrelveld (202), Edds (49), and Hoffman (72) provided anatomic and physiologic evidence that intact fibers from partially cut peripheral nerves can generate axon sprouts that renervate muscles. These studies were followed by Murray and Thompson's (146) 1957 report showing that axonal sprouting also occurred in partially denervated sympathetic ganglion, and Liu and Chambers's (95) 1958 paper indicating that after dorsal spinal root transection there was subsequent axonal sprouting of caudal nerve root fibers into the dorsal spinal cord.

Within a short time other studies demonstrated injury-induced axonal sprouting in the central nervous system (CNS) (66, 158, 167). Anatomically, one of the most convincing was Raisman's 1969 study, which used light and electron microscopy (159). In his experiment, the medial forebrain bundle and/or the fimbria were sectioned, and several weeks later septal neurons were examined for signs of axon degeneration and renervation. After a fimbria lesion, the medial forebrain fibers extended axonal branches into zones once occupied by prior fimbria axons. Likewise, after medial forebrain lesions, fimbria fibers occupied terminals closer to the septal somata that were presumably vacated by medial forebrain fibers. Raisman concluded that after partial nerve fiber injury, deafferented synapses on septal neurons were reoccupied, in time, by surviving axonal branches.

Since these early reports there have been numerous experimental and human studies on the CNS showing injury-induced reactive synaptogenesis (34, 105). This chapter focuses on reactive synaptogenesis and epileptogenesis as it pertains to denervation after status epilepticus (SE) in the hippocampus. The goals are to (1) review the general principles of reactive synaptogenesis in the CNS, (2) discuss the anatomic and physiologic findings of SE-induced hippocampal neuronal loss and axonal sprouting, (3) present experimental evidence supporting the view that “new” aberrant excitatory and inhibitory axon fibers form functional and probably abnormal circuit connections that contribute to spontaneous recurrent seizures, and (4) present the possible molecular mechanisms that may govern the initiation and regulation of SE-induced aberrant axon growth.