Modulation of the central chemoreflex magnitude by the peripheral chemoreceptors: a hyperadditive effect or are we barking up the wrong tree?



Ventilation is required to precisely maintain arterial blood gases and acid–base balance within normal limits. Matching of breathing to metabolic demand involves complex integration of chemically and mechanically mediated afferent feedback to the respiratory control centre in the brain stem. At rest, ventilation is primarily controlled via detection of perturbations in arterial levels of CO2, O2 and H+ by two distinct chemoreceptive elements situated peripherally (primarily the carotid bodies, CB) and centrally (medulla oblongata). Although it is commonly accepted that both the peripheral and central chemoreceptors are important mediators of ventilatory responsiveness, it remains unclear how these receptors interact to provide coordinated control of ventilation.

Previous investigations have reported that the interaction between the peripheral and central chemoreceptive compartments can be (1) additive, where the peripheral and central chemoreceptors do not interact in a significant way, (2) hyperadditive, where activation of one receptor augments the sensitivity or gain of the other, or (3) hypoadditive, where increased stimulation of one receptor attenuates the sensitivity or gain of the other. It has been recently discovered that CB afferents converge on chemosensitive neurons within the medulla and discharge from these afferents appears to impact upon the central respiratory control centre (Takakura et al. 2006). Identification of this neural circuitry has provided evidence of an intriguing link through which the peripheral and central chemoreceptors may interact to regulate ventilation in a manner that is other than simply additive. However, owing to the aforementioned divergence amongst previous findings, exactly how central–peripheral integration controls ventilation remains controversial.

In a study published recently in The Journal of Physiology, Blain et al. (2010) attempted to resolve some of this controversy by evaluating the contribution of CB afferent input to the ventilatory response to CNS hypercapnia in non-anaesthetized, awake dogs. Specifically, the authors tested the functional significance of the recently described neural circuitry that connects the peripheral to the central chemoreceptors by exploiting a novel, complex surgical preparation that allowed independent control of peripheral and central chemoreceptor stimulation/inhibition. After denervation of the left CB, the right carotid sinus was instrumented with a vascular occluder and catheter allowing reversible isolation and extracorporeal perfusion of the remaining intact CB. The extracorporeal circuit was then used to expose the remaining CB to three conditions in a random order: (1) CB normal, perfusate with eupnoeic inline image, inline image and H+ levels; (2) CB inhibition, hyperoxic hypocapnic perfusate; and (3) CB stimulation, hypoxic normocapnic perfusate. In each condition, the authors assessed the alterations in ventilation, diaphragm EMG and blood gases in response to progressive hypercapnia of the systemic circulation, and therefore the CNS and central chemoreceptors, via progressive increases in fractional inspired CO2.

The primary finding of Blain et al. was that both CB inhibition and CB stimulation had profound effects on the magnitude of the central chemoreflex, and thus the ventilatory, response to CNS hypercapnia. CB inhibition decreased the slope of the ventilatory response to progressive CNS hypercapnia, and CB stimulation drastically increased the slope of the ventilatory response to CNS hypercapnia in all dogs. The increased ventilatory slope with CB stimulation was mediated primarily by an increase in breathing frequency and to a lesser extent an increase in tidal volume. Conversely, the reduction in ventilation with CB inhibition was due primarily to a reduction in tidal volume with less contribution from a reduction in breathing frequency. The increase in ventilation with CB stimulation was associated with an increase in diaphragm EMG, whereas the hypoventilation that occurred with CB inhibition was associated with a reduction in diaphragm EMG. Based on these findings, the authors concluded that the sensitivity of the central chemoreflex is highly dependent on CB afferent input and that this peripheral–central interaction is hyperadditive in nature.

The synergistic hyperadditive control of ventilation mediated by peripheral–central integration evidenced by Blain and co-authors differs from many previous reports in the literature. For example, in contrast to the experimental design implemented by Blain et al., Day & Wilson (2009) examined the effect of central chemoreflex stimulation and inhibition on the magnitude of the ventilatory response to perturbations in CB inline image and inline image above as well as below eupnoeic values in non-anaesthetized, decerebrate rats. The investigators demonstrated that the gain of the peripheral chemoreflex was augmented with inhibition of the central chemoreceptors regardless of whether peripheral chemoreceptor activation was increased or decreased, indicating a hypoadditive interaction (Day & Wilson 2009). That is, the efficacy of the peripheral CB chemoreflex in contributing to ventilation was decreased with central chemoreceptor stimulation. With regard to a potential explanation for these contrasting findings, Blain et al. note that decerebration in the study by Day and Wilson may have blunted the ventilatory response to increases in brainstem inline image secondary to removal of chemosensitive suprapontine structures (such as the hypothalamus). However, whether removal of such structures would be sufficient to produce these contrasting findings is unclear. It would be interesting to determine whether central chemoreceptor inhibition/stimulation in the experimental model described by Blain et al. elicited a hypoadditive or hyperadditive effect on the ventilatory response to CB stimulation. If, in agreement with the findings of Day and Wilson, central chemoreceptor stimulation did indeed facilitate a hypoadditive response, then it could be proposed that while the gain of the central chemoreflex can be strongly influenced by CB afferent activity, ventilation is largely dictated by central rather than peripheral chemosensitivity when concurrent stimulation of the central and peripheral chemoreceptors occurs.

There is developing evidence that identifies a number of blood-borne factors other than O2, CO2 and H+ that can inhibit/stimulate CB activity (Kumar & Bin-Jaliah 2007). Indeed the CBs are now considered ‘multi-modal’ physiological sensors that detect and discharge in response to hypoglycaemia, increased temperature, hyperosmolarity, and perturbations in arterial potassium (Kumar & Bin-Jaliah 2007). However, whether CB stimulation via one or more of these alternative physiological stimuli also increases the gain of the central chemoreflex in a manner similar to that described by Blain et al. is unknown and warrants further investigation. Moreover, it would be of great interest to determine whether concurrent CB exposure to multiple stimulators (e.g. hypoxic hypoglycaemia) would further increase the central chemoreceptor response to hypercapnia.

Blain et al. provide excellent evidence that the sensitivity of the central chemoreceptors to CNS hypercapnia is critically dependent on CB afferent activity and that synergistic interaction between the peripheral and central chemoreceptors is hyperadditive in nature. Their interesting report leads to a compelling next question: do the peripheral and central chemoreceptors exhibit a similar synergistic control of ventilation in humans?

In contrast to the findings of Blain et al., the majority of experimental evidence from human studies has suggested that the peripheral and central chemoreceptors do not interact in a significant fashion. For example, Clement et al. (1995) examined the effect of central chemoreflex stimulation (via hypercapnia) on the ventilatory response to peripheral chemoreceptor stimulation through perturbations in arterial pH (pHa) and inline image (normoxia, hypoxia and hyperoxia). Although the slope of the ventilatory response to increasing pHa was attenuated at each level of arterial inline image with augmented central chemoreceptor stimulation, suggesting a hypoadditive interaction, neither the hypoxic sensitivity of the ventilation–pHa slope (i.e. the change in slope between hypoxia and normoxia or hyperoxia) nor the hypoxic sensitivity at a fixed pHa was different in hypercapnia versus normocapnia. Based on these observations the authors concluded that the peripheral and central chemoreflexes independently control breathing in humans.

More recent evidence, however, has suggested that interaction of the peripheral and central chemoreceptors may in fact elicit a synergistic hyperadditive effect on ventilation in humans (Duffin 2010). If it is assumed that ventilation is determined by summation of the peripheral and central chemoreceptor drives in the respiratory control centre, then the ventilatory response to progressive hypercapnia (i.e. rebreathing) should be linear in nature. However, in one of four examples examined, Duffin (2010) demonstrated that the ventilatory response to rebreathing was curvilinear and was likely to be due to a highly responsive peripheral chemoreflex. This observation suggests that, in agreement with Blain et al., peripheral chemoreceptor stimulation may indeed modulate central chemoreceptor sensitivity in humans. We must, however, urge caution when interpreting such results from human studies due to the difficulty in designing and implementing such research, particularly given that the stimuli that are experimentally manipulated are not entirely specific for each set of chemoreceptors.

If the findings of Blain et al. are transferable into human physiology, it could call into question our current understanding of the control of breathing, and in addition may have large implications in clinical populations with regard to disease pathophysiology. It is well known that chemosensitivity is heightened in a number of disease states, including hypertension, obstructive sleep apnoea and heart failure. In heart failure, the up-regulation of the chemoreceptors is thought to be an initial compensatory mechanism to maintain cardiac output, blood pressure and homeostasis. However, as heart failure progresses, over-stimulation of peripheral and central chemoreceptors may have deleterious effects resulting in autonomic imbalance, sympathetic overactivity and increased ventilation and breathlessness. In view of the findings of Blain et al., it is plausible that the chaotic breathing and hyperventilation associated with heart failure may arise from an increase in the gain of the central chemoreceptors secondary to an overactivity of the peripheral chemoreceptors, which is mediated by local changes (i.e. reduced blood flow and reduced blood O2 tension).

In conclusion, Blain and colleagues have demonstrated that central chemoreceptor sensitivity to hypercapnia in the non-anaesthetized intact dog is critically dependent on CB afferent activity. Moreover, this peripheral–central interaction elicits a synergistic hyperadditive ventilatory response to central hypercapnia. Despite these novel findings, the complex interactions between the peripheral and central chemoreceptors and their respective impacts upon autonomic and cardiorespiratory control remains controversial, with many questions yet to be answered. Further work in an intact and integrated system similar to that used in Blain et al. will provide important advances in our understanding of these complex interactions.



The authors thank Drs Jill Barnes, Nisha Charkoudian, Michael Joyner, Thomas Olson and Bruce Johnson for their critical evaluation of this manuscript and helpful suggestions. The authors apologize for not citing all relevant articles due to reference limitations.