The initial stage of auditory signal processing includes the dorsal cochlear nucleus (DCN), which contains many different cell types organized into a layered structure. There are fundamentally three layers, the superficial layer, the fusiform cell layer, and the deep layer, although in some species, the latter can be further subdivided (Brawer et al. 1974; Lorente de Nó 1981; Osen 1969).
The descending branch of the auditory nerve (AN) terminates in the deep layer and endows the DCN with a tonotopic organization (Rose et al. 1959; Ryugo and May 1993). The deep layer contains a variety of multipolar cells, including giant cells that project to the contralateral inferior colliculus (Adams and Warr 1976) and contralateral cochlear nucleus (Cant and Gaston 1982). Vertical cells are also located in the deep layer, and are an important source of inhibition within the DCN as well as both subdivisions of the ventral cochlear nucleus (VCN) (Rhode 1999; Saint-Marie et al. 1991; Voigt and Young 1980, 1990; Zhang and Oertel 1993a).
The superficial layer lies just beneath the ependymal surface of the DCN and contains a mixture of cells including cartwheel, stellate, and Golgi cells (Mugnaini et al. 1980a; Wouterlood and Mugnaini 1984; Wouterlood et al. 1984). The somata of granule cells are distributed throughout the DCN, and their axons form a network of parallel fibers, arranged orthogonally to the tonotopic axis, that serves as the principal input to the superficial layer (Mugnaini et al. 1980b). The granule cell domains receive input from a variety of sources, including the somatosensory system (Itoh et al. 1987; Weinberg and Rustioni 1987; Wright and Ryugo 1996), the vestibular system (Burian and Gstoettner 1988; Kevetter and Perachio 1989), and the descending auditory system (Benson and Brown 1990; Weedman and Ryugo 1996).
Sandwiched between the superficial and deep layers is the fusiform cell layer, comprising an irregular arrangement of the fusiform cells for which it is named. These neurons have relatively large somata and two dendritic arbors, an apical arbor that extends into the superficial layer and a basal arbor that descends into the deep layer (Brawer et al. 1974; Lorente de Nó 1981). The dendritic arbors are flattened in a plane parallel to the DCN isofrequency laminae (Blackstad et al. 1984). Their axons project out of the DCN to the contralateral inferior colliculus by way of the dorsal acoustic stria (DAS) (Adams and Warr 1976).
The apical arbor is densely branched and covered with spines that are the site of excitatory input from parallel fiber axons (Mugnaini et al. 1980a). The basal dendrites are sparser and free of spines. The distal portions of the basal dendrites receive excitatory synapses from AN fibers (Smith and Rhode 1985). Inhibitory inputs to fusiform cells are located primarily on the soma and proximal dendrites and originate from vertical cells (Saint-Marie et al. 1991; Zhang and Oertel 1993c), cartwheel cells (Berrebi and Mugnaini 1991; Zhang and Oertel 1993a), and possibly from stellate cells of the posteroventral cochlear nucleus (PVCN) (Oertel et al. 1990).
Fusiform cells are of particular interest because they represent the majority of the output fibers of the DCN (Adams 1976) and because their bipolar dendritic structure allows them to integrate activity from both of the other DCN layers. It has been suggested, for example, that they are sensitive to the spectral filtering properties of the pinna and that the circuitry of the superficial layer may serve to account for pinna movement (Parsons et al. 2001; Rice et al. 1992; Young et al. 1992, 1995).
Fusiform cells comprise a heterogeneous population in terms of physiological behavior. Response properties in the unanesthetized, decerebrate preparation have been traditionally classified using the response map scheme of Evans and Nelson (1973), as modified and extended by later investigators (Davis et al. 1996; Shofner and Young 1985; Spirou and Young 1991; Young and Voigt 1982). Extracellular recordings in conjunction with antidromic stimulation of the DAS have associated fusiform cells with type III and type IV units in the decerebrate cat (Young 1980). Fusiform cells labeled as part of intracellular recording experiments in the decerebrate gerbil displayed type I/III, type III or type IV-T unit response properties, but never the responses of a classical type IV unit (Ding et al. 1999).
In barbiturate-anesthetized animals, it has been more common to use poststimulus time histograms (PSTHs) for classifying DCN response types (Pfeiffer 1966). Extracellular recordings from the fusiform cell layer were predominantly pauser or buildup units in the anesthetized cat, suggesting that these may be the responses of fusiform cells (Godfrey et al. 1975). This conclusion was confirmed using intracellular recording and labeling techniques (Rhode et al. 1983). Furthermore, Rhode et al. demonstrated that a single fusiform cell can show pauser, chopper, or onset patterns, depending on the frequency and level of the stimulus and on other experimental conditions. Similarly, fusiform cells in vitro respond to depolarizing current pulses with buildup, pauser, or chopper patterns depending on the level of a hyperpolarizing prepulse (Kanold and Manis 1999; Manis 1990).
This report describes a study in which intracellular recording and labeling techniques were used to investigate the response properties of fusiform cells in barbiturate-anesthetized Mongolian gerbils. One aim was to compare the results of this study to the physiology and anatomy of fusiform cells previously reported in the barbiturate-anesthetized cat (Rhode et al. 1983; Rhode and Smith 1986; Smith and Rhode 1985). Such comparison is of particular importance since there are significant differences between decerebrate gerbils and decerebrate cats (Davis et al. 1996). In particular, whereas type IV units comprise 32% of the units encountered in the decerebrate cat (Shofner and Young 1985), they represent only 11% of the units in decerebrate gerbil (Davis et al. 1996). Furthermore, fusiform cells in the gerbil may not have type IV unit properties at all (Ding et al. 1999), while some subset of those in cat almost certainly do (Young 1980). The results qualitatively confirm the earlier studies showing that fusiform cells exhibit a variety of response properties, including pauser, buildup, and chopper, depending on acoustic stimulus and current-clamp parameters (Manis 1990; Rhode et al. 1983).
This work represents part of the doctoral dissertation of K. E. Hancock.