Cardiovascular adjustments during anticipated postural changes.

It is well-documented that feedforward cardiovascular responses occur at the onset of exercise, but it is unclear if such responses are associated with other types of movements. In this study, we tested the hypothesis that feedforward cardiovascular responses occur when a passive (imposed) 60° head-up tilt is anticipated, such that changes in heart rate and carotid artery blood flow (CBF) commence prior to the onset of the rotation. A light cue preceded head-up tilts by 10 sec, and heart rate and CBF were determined for 5-sec time periods prior to and during tilts. Even after these stimuli were provided for thousands of trials spanning several months, no systematic changes in CBF and heart rate occurred prior to tilts, and variability in cardiovascular adjustments during tilt remained substantial over time. We also hypothesized that substitution of 20° for 60° tilts in a subset of trials would result in exaggerated cardiovascular responses (as animals expected 60° tilts), which were not observed. These data suggest that cardiovascular adjustments during passive changes in posture are mainly elicited by feedback mechanisms, and that anticipation of passive head-up tilts does not diminish the likelihood that a decrease in carotid blood flow will occur during the movements.

Neurons in the pontomedullary reticular formation receive converging inputs from the hindlimb and labyrinth.

The integration of inputs from vestibular and proprioceptive sensors within the central nervous system is critical to postural regulation. We recently demonstrated in both decerebrate and conscious cats that labyrinthine and hindlimb inputs converge onto vestibular nucleus neurons. The pontomedullary reticular formation (pmRF) also plays a key role in postural control, and additionally participates in regulating locomotion. Thus, we hypothesized that like vestibular nucleus neurons, pmRF neurons integrate inputs from the limb and labyrinth. To test this hypothesis, we recorded the responses of pmRF neurons to passive ramp-and-hold movements of the hindlimb and to whole-body tilts, in both decerebrate and conscious felines. We found that pmRF neuronal activity was modulated by hindlimb movement in the rostral-caudal plane. Most neurons in both decerebrate (83% of units) and conscious (61% of units) animals encoded both flexion and extension movements of the hindlimb. In addition, hindlimb somatosensory inputs converged with vestibular inputs onto pmRF neurons in both preparations. Pontomedullary reticular formation neurons receiving convergent vestibular and limb inputs likely participate in balance control by governing reticulospinal outflow.

Vestibular nucleus neurons respond to hindlimb movement in the conscious cat.

The limbs constitute the sole interface with the ground during most waking activities in mammalian species; it is therefore expected that somatosensory inputs from the limbs provide important information to the central nervous system for balance control. In the decerebrate cat model, the activity of a subset of neurons in the vestibular nuclei (VN) has been previously shown to be modulated by hindlimb movement. However, decerebration can profoundly alter the effects of sensory inputs on the activity of brain stem neurons, resulting in epiphenomenal responses. Thus, before this study, it was unclear whether and how somatosensory inputs from the limb affected the activity of VN neurons in conscious animals. We recorded brain stem neuronal activity in the conscious cat and characterized the responses of VN neurons to flexion and extension hindlimb movements and to whole body vertical tilts (vestibular stimulation). Among 96 VN neurons whose activity was modulated by vestibular stimulation, the firing rate of 65 neurons (67.7%) was also affected by passive hindlimb movement. VN neurons in conscious cats most commonly encoded hindlimb movement irrespective of the direction of movement (n = 33, 50.8%), in that they responded to all flexion and extension movements of the limb. Other VN neurons overtly encoded information about the direction of hindlimb movement (n = 27, 41.5%), and the remainder had more complex responses. These data confirm that hindlimb somatosensory and vestibular inputs converge onto VN neurons of the conscious cat, suggesting that VN neurons integrate somatosensory inputs from the limbs in computations that affect motor outflow to maintain balance.