RESPONSE TO CO2 OF NEURONS IN THE ROSTRAL VENTRAL MEDULLA IN-VITRO

1. It has been hypothesized that CO2-sensitive neurons an located in t he rostral ventral medulla. To demonstrate this at the cellular level, perforated patch-clamp recordings were made from rat medullary slices in vitro. The effect of respiratory acidosis/alkalosis on the electro physiologic properties of neurons was studied by recording membrane po tential while changing the CO2 of the bath solution and allowing pH to vary. 2. At baseline, most neurons in the rostral ventrolateral medul la (VLM) and rostral medullary raphe spontaneously tired repetitively at a regular rate (3.3 +/- 3.5 Hz, mean +/- SD) with a linear interspi ke ramp depolarization (n = 102 of 135). Spontaneous firing continued after synaptic blockade with high-magnesium, low-calcium solution (n = 14 of 15). Spontaneous firing of calcium spikes continued in tetrodot oxin (TTX; n = 13 of 13), but was blocked by TTX and cadmium (n = 4 of 4). 3. The effect of respiratory acidosis/alkalosis on neurons was ex amined by changing the CO2 of the bicarbonate-buffered bath solution w ithin the range of 3-9%. Most neurons studied (n = 74 of 105) did not change their firing rate in response to this stimulus, however, some n eurons were stimulated (n = 16) and other neurons were inhibited (n = 15) by increases in CO2. 4. In many CO2-stimularcd neurons. the increa se in firing rate caused by an increase in CO2 was associated with an increase in slope of the linear interspike ramp depolarization, wherea s in many CO2-inhibited neurons the opposite occurred, i.e., an increa se in CO2 resulted in a decrease in slope of the ramp depolarization. These changes occurred without a change in the level of afterhyperpola rization polarization or spike threshold. 5. Whole cell patch-clamp re cording invariably resulted in loss of spontaneous and stimulated repe titive firing over 10-40 min despite good resting potential, input res istance, and amplitude of single depolarization-evoked spikes, CO2 pro duced no change in membrane potential in neurons after rundown of repe titive firing. The loss of repetitive firing and CO2 sensitivity with whole cell recording required the use of perforated-patch recordings o f mem brane potential or cell-attached-patch recordings of spike trans ients to accurately study the baseline electrophysiologic properties a nd CO2 sensitivity of rostral medullary neurons. 6. Neuronal location was determined before each recording using direct visualization of liv ing slices, and after some recordings using biocytin staining. CO2-sti mulated and CO2-inhibited neurons were both found to have cell bodies in the rostral VLM, an area thought to contain central respiratory che moreceptors. However, other neurons with similar baseline properties a nd CO2 response were also found in the medullary raphe, an area known to influence respiration but not previously associated with chemorecep tion. 7. These results provide direct evidence for neurons with chemor eceptive properties in VLM. The presence of neurons with a similar res ponse within the medullary raphe contributes to the accumulating evide nce that CO2-sensitive neurons are more widely distributed than previo usly believed. CO2 sensitivity may be a widespread property of neurons within nuclei involved in control or modulation of respiration, or ma y be present in neurons involved in control of other, nonrespiratory, CO2-sensitive CNS functions.