Rishu Bai

Otolith and canal integration on single vestibular neurons in cats

In this review, based primarily on work from our laboratory, but related to previous studies, we summarize what is known about the convergence of vestibular afferent inputs onto single vestibular neurons activated by selective stimulation of individual vestibular nerve branches. Horizontal semicircular canal (HC), anterior semicircular canal (AC), posterior semicircular canal (PC), utricular (UT), and saccular (SAC) nerves were selectively stimulated in decerebrate cats. All recorded neurons were classified as either projection neurons, which consisted of vestibulospinal (VS), vestibulo-oculospinal (VOS), vestibulo-ocular (VO) neurons, or non-projection neurons, which we simply term “vestibular″ (V) neurons. The first three types could be successfully activated antidromically from oculomotor/trochlear nuclei and/or spinal cord, and the last type could not be activated antidromically from either site. A total of 1228 neurons were activated by stimulation of various nerve pair combinations. Convergent neurons were located in the caudoventral part of the lateral, the rostral part of the descending, and the medial vestibular nuclei. Otolith-activated vestibular neurons in the superior vestibular nucleus were extremely rare. A high percentage of neurons received excitatory inputs from two nerve pairs, a small percentage received reciprocal convergent inputs and even fewer received inhibitory inputs from both nerves. More than 30% of vestibular neurons received convergent inputs from vertical semicircular canal/otolith nerve pairs. In contrast, only half as many received convergent inputs from HC/otolith-nerve pairs, implying that convergent input from vertical semicircular canal and otolith-nerve pairs may play a more important role than that played by inputs from horizontal semicircular canal and otolith-nerve pairs. Convergent VS neurons projected through the ipsilateral lateral vestibulospinal tract (i-LVST) and the medial vestibulospinal tract (MVST). Almost all the VOS neurons projected through the MVST. Convergent neurons projecting to the oculomotor/trochlear nuclei were much fewer in number than those projecting to the spinal cord. Some of the convergent neurons that receive both canal and otolith input may contribute to the short-latency pathway of the vestibulocollic reflex. The functional significance of these convergences is discussed.

Convergence of ipsilateral semicircular canal inputs onto single vestibular nucleus neurons in cats.

Convergent inputs from the ipsilateral semicircular canal nerves onto single vestibular nucleus neurons were investigated in decerebrate cats using intracellular recording after selective stimulation of each ampullar nerve. One hundred and seventy-four neurons were activated by stimulating the anterior semicircular (AC) and/or posterior semicircular canal (PC) nerves. These neurons were also antidromically stimulated and classified according to the pattern of their collateral projections to the oculomotor complex and the spinal cord. Four types were found: vestibulo-ocular (VO), vestibulospinal (VS), vestibulo-oculospinal (VOS), and vestibular (V) neurons, the latter of which were not activated by stimulation of either the oculomotor complex or the spinal cord. Of 174 AC- and/or PC-activated vestibular nucleus neurons, 32 (18%) received convergent inputs from both nerves. These convergent neurons included 11 VS, 6 VOS, and 15 V neurons. We found no VO neurons with convergent input. The vast majority (82%) of AC/PC-activated VS and VOS convergent neurons received excitatory inputs from both nerves, 12% received reciprocal inputs (i.e., excitatory from one and inhibitory from the other), and the remaining neurons received inhibitory inputs from both nerves. By stimulating the horizontal semicircular (HC) and/or PC nerves, 183 neurons were activated. Of these, 44 (24%) received convergent inputs from both nerves. These convergent neurons included 19 VS, 5 VOS, 2 VO, and 18 V neurons. Approximately one-half (46%) of HC/PC-activated VS and VOS convergent neurons received excitatory inputs from both nerves and 42% received reciprocal inputs, and the remaining neurons received inhibitory inputs from both nerves. In both nerve pairs, the percentage of VS neurons was higher (AC/PC, 34%; HC/PC, 43%) than that of VOS or VO neurons. Approximately half of these convergent neurons were located in the lateral nucleus. These results suggest that, during mixed angular head accelerations, the vestibulocollic reflex may be partly accomplished by VS and VOS convergent neurons.

Axonal pathways and projection levels of anterior semicircular canal nerve-activated vestibulospinal neurons in cats

Using collision tests of orthodromically and antidromically generated spikes, we studied the axonal pathways, axonal projection levels, and soma location of anterior semicircular canal (AC) nerve-activated vestibulospinal neurons in decerebrate cats. AC nerve-activated vestibulospinal neurons (n=74) were mainly located in the ventral portion of the lateral vestibular nuclei and the rostral portion of the descending vestibular nucleus, which is consistent with previous studies. Of these neurons, 15% projected through the ipsilateral (i-) lateral vestibulospinal tract (LVST), 74% projected through the medial vestibulospinal tract (MVST), and 11% projected through the contralateral (c-) LVST. The vast majority (78%) of AC nerve-activated vestibulospinal neurons were activated antidromically only from the cervical segment of the spinal cord; 15% of neurons were activated from the T1 segment and only one neuron was activated from the L3 segment. AC nerve-activated vestibulospinal neurons may primarily target the neck muscles and thus contribute to the vestibulocollic reflex. Most of the c-LVST neurons were also activated antidromically from the oculomotor nucleus, suggesting that they are closely related to the control of combined eye-head movements.

Properties of horizontal semicircular canal nerve-activated vestibulospinal neurons in cats

Axonal pathways, projection levels, and locations of horizontal semicircular canal (HC) nerve-activated vestibulospinal neurons were studied. The HC nerve was selectively stimulated. Vestibulospinal neurons were activated antidromically with four stimulating electrodes, inserted bilaterally into the lateral vestibulospinal tracts (LVST) and medial vestibulospinal tracts (MVST) at the C1/C2 junction. Stimulating electrodes were also positioned in the C3, T1, and L3 segments and in the oculomotor nuclei. Most HC nerve-activated vestibulospinal neurons were located in the ventral portion of the medial, lateral, and the descending nuclei. Among the 157 HC nerve-activated vestibular neurons, 83 were antidromically activated by stimulation at the C1/C2 junction. Of these 83 neurons, axonal pathways of 56 HC nerve-activated vestibulospinal neurons were determined. Most (48/56) of these had axons that descended through the MVST, with the remainder (8 neurons) having axons that descended through the ipsilateral (i-) LVST. Laterality of the axons’ trajectories through the MVST was investigated. The majority of vestibulospinal neurons (24/28) with axons descending through the contralateral MVST were also antidromically activated from the oculomotor nucleus, whereas almost all vestibulospinal neurons (19/20) with axons descending through the i-MVST were not. Most HC nerve-activated vestibulospinal neurons were activated antidromically only from the C1/C2 or C3 segments. Only one neuron that was antidromically activated from the T1 segment had an axon that descended through the i-LVST. None of the HC nerve-activated vestibulospinal neurons were antidromically activated from the L3 segment. It is likely that the majority of HC nerve-activated vestibulospinal neurons terminate in the cervical cord and have strong connections with neck motoneurons.

Morphology of physiologically identified otolith-related vestibular neurons in cats.

The morphology of physiologically identified otolith nerve-activated vestibular neurons was investigated using intracellular injections of horseradish peroxidase (HRP). Eleven utricular, 11 saccular and three utricular/saccular nerve-activated vestibular neurons were labeled with HRP. All of these neurons except one were secondary neurons, the exception being a convergent neuron. The labeled neurons were pyramidal, elongated and ovoidal in shape. Most of the labeled cells were medium to large (mean diameter: > or =30 micro m). There was no apparent correlation between morphology and the different types of otolith nerve-activated vestibular neurons. Thus, it seems likely that the functional type of vestibular neurons cannot be presumed on the basis of their morphology alone.