Sympathetic activation by the central chemoreceptor ‘reflex’: new evidence that RVLM vasomotor neurons are involved … but are they enough?

The existence of chemoreceptive neurons in the central nervous system was discovered more than 40 years ago (Mitchell et al. 1963). Subsequently, studies have demonstrated that cerebrospinal fluid (CSF) acidification by even a small increase in dissolved carbon dioxide (CO₂) (i.e. hypercapnoea) activates chemoreceptor neurons and thereby stimulates the ‘drive to breathe’ (Feldman et al. 2003; Nattie & Li, 2006). Whether respiratory stimulation is mediated exclusively/predominantly by neurons within the ventral surface of the medulla as suggested by early (Mitchell et al. 1963) and more recent (Mulkey et al. 2004) studies, or is an emergent property of a brainstem network of chemoreceptor neurons (Smith et al. 2006) continues to be a source of spirited dialog (Nattie & Li, 2006). In this issue of The Journal of Physiology, Moreira et al. (2006) shift attention toward another important component of the response to hypercapnoea, namely increased sympathetic nerve discharge (SND). Though less well studied than respiratory effects of hypercapnoea, sympathetic activation by elevated arterial/CSF CO₂ has important homeostatic functions that maintain appropriate cerebral and systemic blood flow distribution during respiratory depression, when inspired gases become hypercapnic, or when acid–base balance is disrupted (Dean et al. 1990; Feldman et al. 2003; Smith et al. 2006).

Using electrophysiological recording methods in vivo, Moreira et al. (2006) demonstrate that reticulospinal vasomotor neurons in the rostral ventrolateral medulla (RVLM) increase their firing rate when end expiratory CO₂ (an index of arterial P_CO2) is increased from 5 to 10%. Responsive cells were shown to consist of relatively fast-conducting, spinal chord-projecting C1 adrenergic (phenylethanolamine N-methyl transferase-positive) and non-C1 cells within the vasomotor control region of the RVLM. Autonomic neurophysiologists are keenly aware that sympathoexcitatory neurons of the RVLM play pivotal roles in the generation of basal SND and its modulation by humoral factors (e.g. blood-borne ANG II, plasma hyperosmolality, etc.) (Toney et al. 2003) and visceral afferent reflexes (e.g. the arterial baroreflex, arterial chemoreflex, somato-sympathetic reflex, etc.) (Madden & Sved, 2003). By documenting involvement of RVLM sympathoexcitatory neurons in hypercapnoea-induced activation of SND, the present study adds to an already extensive list of functions subserved by these important vasomotor control neurons.

It is well recognized that ongoing SND is modulated by respiration and that aspects of this ‘respiratory entrainment’ are maintained in the absence of arterial chemoreceptor and cardio-respiratory vagal afferent inputs. Neurons in the pre-Bötzinger/rostral ventral respiratory group (CVLM region, part of the central respiratory pattern generator network) appear to contribute significantly to ‘central’ respiratory modulation of SND (Feldman et al. 2003). The mechanism appears to depend, at least in part, on phasic inhibition of RVLM sympathoexcitatory neurons by GABAergic neurons in the CVLM. Using microinjection methods, the study by Moreira et al. (2006) indicates that sympathoexcitation by hypercapnoea does not depend on inputs from the CVLM, rather CVLM inputs (presumably to the RVLM) attenuate hypercapnoea-induced sympathoexcitation. Likewise, SND responses to elevated end expiratory CO₂ do not require input from the caudal NTS, a dorsal medullary region reported to contain neurons with intrinsic CO₂ sensitivity (Dean et al. 1990).

Placed in perspective, the study by Moreira et al. (2006) indicates that RVLM sympathoexcitatory neurons respond to hypercapnoea in anaesthetized rats after removal of cardio-respiratory visceral afferent inputs. It is presently uncertain whether this is due to their possible intrinsic sensitivity to CSF acidification by elevated CO₂ (Washburn et al. 2003), to synaptic activation by one or more groups of CO₂-sensitive neurons (Nattie & Li, 2006), or both. Whatever the case might be, the present data argue against a synaptic mechanism involving glutamate, as microinjection into the RVLM of the ionotropic excitatory amino acid receptor antagonist kynurenate failed to attenuate the increase in SND induced by hypercapnoea. The latter point is both interesting and perhaps surprising because it indicates that nearby glutamatergic neurons in the retrotrapezoid nucleus, which were earlier shown by the same laboratory to be highly sensitive to elevated CO₂ (Mulkey et al. 2004) are not required.

The work of Moreira et al. (2006) emphasizes the value of applying multiple experimental approaches (in vivo single neuron recording, neurochemical phenotyping, microinjection techniques, haemodynamic and peripheral nerve recordings) in a single neurophysiological study. Investigators will and should continue to study central autonomic regulation in vivo by exploring stimulus specific mechanisms that modulate the activity and excitability of identified sympathetic regulatory neurons. These mechanisms will probably be more completely delineated by consistently recognizing that RVLM neurons (and potentially other sympathetic control neurons) are governed by respiratory influences of both peripheral AND central origin. The work of Moreira et al. demonstrates how important it is to document such convergent/concurrent influences.