Glucose, insulin, and the carotid body chemoreceptors in humans

Abstract

Known primarily for its oxygen-sensing capabilities, the carotid body chemoreceptors have recently been implicated, primarily by work in animal models, in the pathophysiology of a number of metabolic conditions. The research presented in this brief review highlights translational work conducted at the Mayo Clinic between 2010 and 2017 in healthy humans and discusses key areas for future work in disease populations.

INTRODUCTION

The purpose of this review is to highlight a presentation given on August 4, 2017 at the International Union of Physiological Sciences conference in Rio de Janeiro, Brazil, in a symposium entitled “Polymodal properties of the carotid body chemoreceptors beyond hypoxia.” Known primarily for their oxygen-sensing capabilities, the carotid body chemoreceptors have recently been implicated, primarily by work in animal models, in the pathophysiology of a number of metabolic conditions. The research presented in this review highlights translational work conducted at the Mayo Clinic between 2010 and 2017 in healthy humans (15, 21, 25, 26, 31, 51, 52) and discusses key areas for future work in disease populations.

STUDYING CAROTID BODY CHEMORECEPTOR FUNCTION IN HUMANS

The carotid body chemoreceptors are located bilaterally at the bifurcation of each common carotid artery. The main purpose of the carotid chemoreceptors is to detect chemical changes within the peripheral arterial blood and initiate reflex changes in ventilation and sympathetic nervous system activity, with the primary known stimulus being a reduction in the partial pressure of oxygen (19, 38). Indeed, healthy humans exhibit robust increases in ventilation in response to hypoxia, termed the “hypoxic ventilatory response.” Assessed as the slope of the relationship between oxygen saturation and minute ventilation, the hypoxic ventilatory response is often determined by exposing individuals to a few short breaths of low (0.05 $F{I}_{O_2}$) or no (0.00 $F{I}_{O_2}$) oxygen and measuring reflex changes in ventilation (3, 11, 12). Although the method is quite repeatable, our group and others have shown there to be large variability in the hypoxic ventilatory response in healthy humans (23, 33). Despite this, there are clear increases in the hypoxic ventilatory response in human disease states, including those related to metabolic diseases (5, 36). In addition, our group and others have shown individuals who have had their carotid bodies resected have essentially no ventilatory response to hypoxia (21, 26, 45, 46, 52) (Fig. 1, A and B), supporting the effectiveness of testing the hypoxic ventilatory response as a measure of carotid body chemosensitivity.

Fig. 1.

Hypoxic ventilatory response is attenuated in individuals who have undergone bilateral carotid body resection. Carotid body chemosensitivity (hypoxic ventilatory response) is blunted in carotid body resected patients (n = 5) when compared with healthy controls (n = 4). *P < 0.01 vs. Control. Figure originally presented in (52).

Independently of completely removing the carotid bodies, a role for the carotid chemoreceptors under resting/basal conditions and in response to physical and/or environmental stress in conscious humans can be assessed via methodologies such as acute hyperoxia [0.40–1.0 $F{I}_{O_2}$ (8, 54)] or intravenous low-dose dopamine (23, 33, 53). Prolonged exposure to high levels of inspired oxygen can elicit reductions in peripheral blood flow, increases in arterial blood pressure, and tachycardia. However, when hyperoxia is employed experimentally in the form of a modified Dejours test (<2 min of 100% oxygen), it has been shown to be a reliable way to assess and/or attenuate peripheral arterial chemosensitivity (7, 8). This is especially true when other variables such as respiratory rate, tidal volume, and/or end-tidal carbon dioxide are also carefully controlled (44). With this, the utility of experimental hyperoxia is limited in situations where long-term changes in blood pressure, blood flow, or responses to hypoxia are of interest. In such cases, low-dose dopamine may be a more useful methodological approach to decrease afferent activity of the carotid body chemoreceptors in humans. Dopamine in low doses (<5 µg/kg/min) is thought to blunt carotid sinus nerve discharge by binding to dopaminergic (primarily D₂) receptors on carotid body Type 1 glomus cells to block calcium currents, leading to membrane hyperpolarization, reduced neurotransmitter release, and decreased carotid sinus nerve discharge (1, 2, 4, 13, 20, 27, 42). However, the independent effects of exogenous intravenous dopamine on the carotid body chemoreceptors may be dose dependent, examined in detail by our group (23). Thus a careful dosing scheme may be necessary to reliably use dopamine as an experimental tool to examine carotid body activity in humans. It is also important to acknowledge that exogenous dopamine, even in low doses, could increase circulating glucagon and/or growth hormone (15, 41, 48, 55), both which may confound studies interested in glucose regulation.

GLUCOSE REGULATION AND THE CAROTID CHEMORECEPTORS

Intermittent hypoxia, blood glucose, and contributing mechanisms.

Anecdotal evidence, available for many years, suggests the carotid body chemoreceptors, key oxygen-sensing sensors, may also play a role in glucose regulation (19). For example, 30 min of continuous hypoxia results in glucose intolerance in healthy humans (34). Furthermore, individuals exposed repeatedly to hypoxia during sleep (apnea hypopnea index >10 events per hour) exhibit high fasting glucose and an impaired glucose-insulin ratio (30). Along these lines, blood glucose is increased during intermittent hypoxia in rats (39), and fasting hyperglycemia observed following chronic intermittent hypoxia can be prevented by carotid body denervation (43). Together, these discoveries led our group to examine the effects of experimental intermittent hypoxia on blood glucose levels in humans. Previous work had suggested 8 h of intermittent hypoxia resulted in increases in glucose and impairments in insulin sensitivity (28). However, such prolonged exposures were unlikely to have isolated effects on the carotid chemoreceptors. Therefore, we sought to examine the effect of a shorter hypoxic exposure on blood glucose levels in healthy humans and the potential role of the carotid body chemoreceptors. To our surprise, glucose measured in venous blood was increased after only 30 min of exposure to intermittent hypoxia (~12 events, 8–10% reduction in S_pO₂ separated by 2 min normoxia) in healthy humans and remained high for 3 h (Fig. 2) (31). Furthermore, this increase in blood glucose occurred independently of any changes in insulin sensitivity (31). We also a observed a positive relationship, albeit weak, between glucose and the hypoxic ventilatory response, suggesting individuals with the greatest increase in carotid chemosensitivity following intermittent hypoxia had the highest blood glucose (32). Although intriguing, this study leaves more questions to be answered, such as: 1) where is this glucose coming from (changes in peripheral blood flow and glucose distribution? changes in insulin secretion? increased gluconeogenesis?), 2) are these effects exacerbated in individuals with sleep apnea and/or diabetes, 3) can this increase in glucose with intermittent hypoxia be prevented (exercise? medication?), and lastly, 4) what is the role of the carotid chemoreceptors in this response?

Fig. 2.

Venous blood glucose during 3 h of intermittent hypoxia or continuous normoxia in healthy humans. Glucose levels remained unchanged during the control visit (P > 0.05 for all). During the intermittent hypoxia visit, glucose levels were greater than T₀ at all time points (*P < 0.01 for all). When data were compared between visits, there was a significant difference at T₃₀ (†P = 0.02) and T₁₈₀ (†P = 0.03). Figure originally presented in (31).

Hypoglycemia counterregulation: acute responses vs. long-term adaptations.

In contrast to increases blood glucose with hypoxia, others have found oxygen therapy improve glucose tolerance and/or insulin sensitivity in patients with diabetes and chronic obstructive pulmonary disease (10, 14, 49). Although contributing mechanisms are unclear, some have suggested that hyperoxia works by attenuating the ability of key peripheral receptors (i.e., carotid chemoreceptors) to sense and/or respond to alterations in blood glucose (49). In line with this idea, Pardal and Lopez-Barneo (35) showed that the carotid bodies exhibit a secretory response during exposure to low glucose. Furthermore, removal of the carotid body chemoreceptors in dogs impaired their ability to respond to hypoglycemia and initiate appropriate counterregulatory responses (18). With this information in mind, we sought to examine whether decreases in carotid chemoreceptor activity would decrease mobilization of key glucoregulatory (e.g., glucagon, growth hormone, epinephrine, cortisol) hormones and impair normal glucose control in healthy humans.

Using systemic hyperoxia to attenuate carotid body afferent activity, Wehrwein and colleagues (51) showed the normal rise in glucoregulatory hormones (e.g., epinephrine, cortisol, glucagon, growth hormone) during a hyperinsulinemic-hypoglycemic clamp was impaired in healthy adult humans. These data (51) were some of the first in humans to show by acutely reducing carotid body chemoreceptor activity the counterregulatory hormone response to hypoglycemia is attenuated. However, the chronic effect of carotid chemoreceptor “desensitization” on glucose control was unclear. For this reason, we recruited five individuals who had their carotid bodies resected bilaterally (for the removal of glomus tumors) to complete the same experiment (52). These individuals underwent two hyperinsulinemic-hypoglycemic clamps during normoxia and hyperoxia, completed on separate days. As we expected, there was no observable effect of hyperoxia on hypoglycemia counterregulation in individuals who had their carotid body chemoreceptors removed, suggesting any effect of hyperoxia on glucose regulation in healthy adults was via the carotid body chemoreceptors (52).

We also found, quite unexpectedly, individuals who had their carotid bodies resected exhibited essentially normal glucoregulatory responses to hypoglycemia under normoxic conditions (52). At first glance these data would suggest long-term removal of the carotid body chemoreceptors does not impact normal glucose control. However, after closer examination of the data we observed two key relationships in the individuals studied. First, there was an association between the time from bilateral resection (years) and the presence of carotid chemoreceptor sensitivity (hypoxic ventilatory response) such that those individuals that underwent surgery most recently exhibited the greatest impairments in chemoreceptor sensitivity to hypoxia (Fig. 3A). Second, those patients with the lowest level of chemoreceptor sensitivity tended to require the greatest amount of glucose infused during the hyperinsulinemic-hypoglycemic clamp (Fig. 3B). In other words, the level of impairment in the counterregulatory response to hypoglycemia was related to the level of residual carotid body chemosensitivity. Together these data suggest the carotid body chemoreceptors play an important role in the counterregulatory response to hypoglycemia when examined acutely in healthy humans and/or shortly after bilateral carotid body resection. However, long-term adaptations likely occur following carotid body resection in humans such that redundant mechanisms are able to recover any previous role of the carotid body chemoreceptors in glucose regulation.

Fig. 3.

Relationship between carotid body chemosensitivity and time from surgery (A), hypoglycemia counterregulation (B). A: those patients that underwent carotid body resection most recently exhibited the greatest impairments in chemoreceptor sensitivity to hypoxia (r = −0.835, P = 0.08). B: those patients with the lowest level of chemoreceptor sensitivity (hypoxic ventilatory response) required the greatest amount of glucose infused during the hyperinsulinemic-hypoglycemic clamp (r = 0.798, P = 0.11). Figure originally presented in (52).

POTENTIALLY CONFOUNDING EFFECTS OF INSULIN

Now until this time, the majority of our experiments designed to access a role for the carotid chemoreceptors in glucose regulation in humans occurred in the setting of high insulin (e.g., hyperinsulinemic clamp). However, in the midst of our work Ribeiro and colleagues (40) found insulin receptors on the carotid bodies and showed insulin increases carotid body neurosecretion and elicits hyperventilation in rats. Because high levels of insulin may have confounded our previous findings (22), we followed up our previous work by asking two important questions: 1) Can we experimentally decrease blood glucose in humans independently of insulin? and 2) Do the carotid bodies sense and/or respond to insulin in humans?

Decreasing blood glucose independently of insulin.

Koyama and colleagues (17) showed in dogs that prolonged high-intensity exercise could elicit a significant reduction in glucose independently of insulin (50% V̇o_2max for 150 min, 6 mg/dl reduction in glucose by ~90 min). Furthermore, in the same model of exercise, carotid body resection resulted in fall in blood glucose that was more severe (~11 mg/dl), occurred much earlier (within 30 min), and glucose remained low throughout the entire exercise protocol (17). Thus, to further our understanding of the role for the carotid bodies in glucose regulation and translate these findings to humans, our group, led by Blair Johnson (15), recruited elite cyclists known for their ability to exercise at a high intensity for a long period of time. These individuals completed an upright cycling protocol at 65% peak oxygen consumption for up to 120 min under a control (saline) condition and during intravenous infusion of low-dose dopamine. Similar to what was seen in dogs, decreasing carotid body afferent activity with low-dose dopamine resulted in a greater fall in blood glucose during the exercise protocol in the elite cyclists studied. As a result, exercise duration was also significantly reduced during dopamine infusion when compared with control (15). These data support a role for the carotid chemoreceptors in normal glucose control during prolonged exercise in healthy exercise-trained individuals and are agreement with our earlier work using hyperinsulinemic-hypoglycemic clamps. Thus, using an insulin-independent model of hypoglycemia, we were able to confirm previous work suggesting the carotid chemoreceptors play a role in normal glucose control.

Insulin-sensing capabilities of the carotid chemoreceptors in humans.

On the basis of the data presented above, it appears the carotid body chemoreceptors play at least a minor role in glucose regulation in healthy humans. However, in day-to-day situations, it is rare to observe changes in glucose and/or insulin independent of the other. With this, we continued to ask the question: What is the role of the carotid body chemoreceptors in the reflex response to insulin? As noted above, Ribeiro and colleagues (40) found insulin receptors on the carotid bodies and showed insulin increases carotid body neurosecretion and elicits hyperventilation in rats. Furthermore, in humans, the hypoxic ventilatory response has been shown to be increased in the presence of insulin in the form of a hyperinsulinemic clamp (50) and in response to a meal (56). It is commonly accepted that even small increases in plasma insulin concentrations have marked sympathoexcitatory effects, and data from animals have shown injection of insulin into the carotid artery increases blood pressure through a sympathetic mechanism (37). Therefore, we sought to explore a potential role for the carotid body chemoreceptors in insulin-mediated sympathoexcitation. To do this, we completed hyperinsulinemic, euglycemic clamps in healthy humans and measured reflex changes in ventilation (tidal volume, respiratory rate) and sympathetic nervous system activity [muscle sympathetic nerve activity via microneurography (47)]. Healthy individuals completed experiments under normoxic and hyperoxic conditions, in addition to saline and low-dose dopamine infusions. As expected, both muscle sympathetic nerve activity and minute ventilation significantly increased in the presence of increases in systemic insulin; however, any effect of acute hyperoxia did not differ between baseline and hyperinsulinemia (25). These data suggest the carotid chemoreceptors do not contribute acutely to insulin-mediated increases in sympathetic nervous system activity in healthy humans (25). However, when low-dose dopamine was given before the hyperinsulinemic clamp, the rise in sympathetic activity with insulin was attenuated (25). Although we observed potentially confounding changes in basal sympathetic activity with exogenous dopamine administration, these data may support a role for the carotid chemoreceptors in more long-term exposure to insulin (25). With this, it would be interesting to see whether similar conclusions would be made if the individuals studied were insulin resistant (5, 6, 22).

NEW QUESTIONS AND FUTURE DIRECTIONS

In summary, we have been able to translate a number of key studies in animals to humans and have expanded upon our understanding of the role for the carotid body chemoreceptors in glucose regulation. By manipulating the level of carotid body activation in healthy humans, we can significantly increase fasting glucose (intermittent hypoxia) and impair normal glucose regulation (acute hyperoxia, low-dose dopamine). Much of the observed changes in glucose control appear to occur independently of changes in insulin; however, a limitation of the current body of knowledge is the completion of studies in only relatively young, healthy adults. Whether similar findings can be translated to patient populations known for exhibiting impairments in glucose control (e.g., insulin resistance, diabetes) has yet to be thoroughly addressed; however, there are strong data to suggest, in disease populations, the role for the carotid bodies in glucose regulation (and/or dysregulation) may be significant (36, 40, 49). With this, there are clear opportunities for long-term adaptation in glucose control following carotid body resection (52). Thus, for carotid body-mediated therapies to become a viable avenue for treatment, it becomes increasingly important to build a better understanding of: 1) what are the carotid bodies sensing (16), 2) who would benefit from such therapies and how can they be identified (24), 3) are there nonpharmacological and/or nonsurgical methods that may be beneficial (9, 29)?

GRANTS

Funding for the work described in this paper came from: DK-090541 (Joyner); HL-83947 (Joyner); HL-130339 (Limberg); AHA15SDG25080095 (Limberg); DK-084624 (Wehrwein).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author.

AUTHOR CONTRIBUTIONS

J.K.L. conceived and designed research, prepared figures, drafted manuscript, edited and revised manuscript, approved final version of manuscript.